Maintained by Difan Deng and Marius Lindauer.
The following list considers papers related to neural architecture search. It is by no means complete. If you miss a paper on the list, please let us know.
Please note that although NAS methods steadily improve, the quality of empirical evaluations in this field are still lagging behind compared to other areas in machine learning, AI and optimization. We would therefore like to share some best practices for empirical evaluations of NAS methods, which we believe will facilitate sustained and measurable progress in the field. If you are interested in a teaser, please read our blog post or directly jump to our checklist.
Transformers have gained increasing popularity in different domains. For a comprehensive list of papers focusing on Neural Architecture Search for Transformer-Based spaces, the awesome-transformer-search repo is all you need.
2025
4364.
Rahman, Abdur; Street, Jason; Wooten, James; Marufuzzaman, Mohammad; Gude, Veera G.; Buchanan, Randy; Wang, Haifeng
MoistNet: Machine vision-based deep learning models for wood chip moisture content measurement Journal Article
In: Expert Systems with Applications, vol. 259, pp. 125363, 2025, ISSN: 0957-4174.
@article{Rahman_2025,
title = {MoistNet: Machine vision-based deep learning models for wood chip moisture content measurement},
author = {Abdur Rahman and Jason Street and James Wooten and Mohammad Marufuzzaman and Veera G. Gude and Randy Buchanan and Haifeng Wang},
url = {http://dx.doi.org/10.1016/j.eswa.2024.125363},
doi = {10.1016/j.eswa.2024.125363},
issn = {0957-4174},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
journal = {Expert Systems with Applications},
volume = {259},
pages = {125363},
publisher = {Elsevier BV},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
4363.
Jiang, Zhiying; Liu, Risheng; Yang, Shuzhou; Zhang, Zengxi; Fan, Xin
DRNet: Learning a dynamic recursion network for chaotic rain streak removal Journal Article
In: Pattern Recognition, vol. 158, pp. 111004, 2025, ISSN: 0031-3203.
@article{JIANG2025111004,
title = {DRNet: Learning a dynamic recursion network for chaotic rain streak removal},
author = {Zhiying Jiang and Risheng Liu and Shuzhou Yang and Zengxi Zhang and Xin Fan},
url = {https://www.sciencedirect.com/science/article/pii/S0031320324007556},
doi = {https://doi.org/10.1016/j.patcog.2024.111004},
issn = {0031-3203},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
journal = {Pattern Recognition},
volume = {158},
pages = {111004},
abstract = {Image deraining refers to removing the visible rain streaks to restore the rain-free scenes. Existing methods rely on manually crafted networks to model the distribution of rain streaks. However, complex scenes disrupt the uniformity of rain streak characteristics assumed in ideal conditions, resulting in rain streaks of varying directions, intensities, and brightness intersecting within the same scene, challenging the deep learning based deraining performance. To address the chaotic rain streak removal, we handle the rain streaks with similar distribution characteristics in the same layer and employ a dynamic recursive mechanism to extract and unveil them progressively. Specifically, we employ neural architecture search to determine the models of different rain streaks. To avoid the loss of texture details associated with overly deep structures, we integrate multi-scale modeling and cross-scale recruitment within the dynamic structure. Considering the application of real-world scenes, we incorporate contrastive training to improve the generalization. Experimental results indicate superior performance in rain streak depiction compared to existing methods. Practical evaluation confirms its effectiveness in object detection and semantic segmentation tasks. Code is available at https://github.com/Jzy2017/DRNet.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Image deraining refers to removing the visible rain streaks to restore the rain-free scenes. Existing methods rely on manually crafted networks to model the distribution of rain streaks. However, complex scenes disrupt the uniformity of rain streak characteristics assumed in ideal conditions, resulting in rain streaks of varying directions, intensities, and brightness intersecting within the same scene, challenging the deep learning based deraining performance. To address the chaotic rain streak removal, we handle the rain streaks with similar distribution characteristics in the same layer and employ a dynamic recursive mechanism to extract and unveil them progressively. Specifically, we employ neural architecture search to determine the models of different rain streaks. To avoid the loss of texture details associated with overly deep structures, we integrate multi-scale modeling and cross-scale recruitment within the dynamic structure. Considering the application of real-world scenes, we incorporate contrastive training to improve the generalization. Experimental results indicate superior performance in rain streak depiction compared to existing methods. Practical evaluation confirms its effectiveness in object detection and semantic segmentation tasks. Code is available at https://github.com/Jzy2017/DRNet.
4362.
Wu, Zhenpeng; Chen, Jiamin; Al-Sabri, Raeed; Oloulade, Babatounde Moctard; Gao, Jianliang
Asymmetric augmented paradigm-based graph neural architecture search Journal Article
In: Information Processing & Management, vol. 62, no. 1, pp. 103897, 2025, ISSN: 0306-4573.
@article{WU2025103897,
title = {Asymmetric augmented paradigm-based graph neural architecture search},
author = {Zhenpeng Wu and Jiamin Chen and Raeed Al-Sabri and Babatounde Moctard Oloulade and Jianliang Gao},
url = {https://www.sciencedirect.com/science/article/pii/S0306457324002565},
doi = {https://doi.org/10.1016/j.ipm.2024.103897},
issn = {0306-4573},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
journal = {Information Processing & Management},
volume = {62},
number = {1},
pages = {103897},
abstract = {In most scenarios of graph-based tasks, graph neural networks (GNNs) are trained end-to-end with labeled samples. Labeling graph samples, a time-consuming and expert-dependent process, leads to huge costs. Graph data augmentations can provide a promising method to expand labeled samples cheaply. However, graph data augmentations will damage the capacity of GNNs to distinguish non-isomorphic graphs during the supervised graph representation learning process. How to utilize graph data augmentations to expand labeled samples while preserving the capacity of GNNs to distinguish non-isomorphic graphs is a challenging research problem. To address the above problem, we abstract a novel asymmetric augmented paradigm in this paper and theoretically prove that it offers a principled approach. The asymmetric augmented paradigm can preserve the capacity of GNNs to distinguish non-isomorphic graphs while utilizing augmented labeled samples to improve the generalization capacity of GNNs. To be specific, the asymmetric augmented paradigm will utilize similar yet distinct asymmetric weights to classify the real sample and augmented sample, respectively. To systemically explore the benefits of asymmetric augmented paradigm under different GNN architectures, rather than studying individual asymmetric augmented GNN (A2GNN) instance, we then develop an auto-search engine called Asymmetric Augmented Graph Neural Architecture Search (A2GNAS) to save human efforts. We empirically validate our asymmetric augmented paradigm on multiple graph classification benchmarks, and demonstrate that representative A2GNN instances automatically discovered by our A2GNAS method achieve state-of-the-art performance compared with competitive baselines. Our codes are available at: https://github.com/csubigdata-Organization/A2GNAS.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
In most scenarios of graph-based tasks, graph neural networks (GNNs) are trained end-to-end with labeled samples. Labeling graph samples, a time-consuming and expert-dependent process, leads to huge costs. Graph data augmentations can provide a promising method to expand labeled samples cheaply. However, graph data augmentations will damage the capacity of GNNs to distinguish non-isomorphic graphs during the supervised graph representation learning process. How to utilize graph data augmentations to expand labeled samples while preserving the capacity of GNNs to distinguish non-isomorphic graphs is a challenging research problem. To address the above problem, we abstract a novel asymmetric augmented paradigm in this paper and theoretically prove that it offers a principled approach. The asymmetric augmented paradigm can preserve the capacity of GNNs to distinguish non-isomorphic graphs while utilizing augmented labeled samples to improve the generalization capacity of GNNs. To be specific, the asymmetric augmented paradigm will utilize similar yet distinct asymmetric weights to classify the real sample and augmented sample, respectively. To systemically explore the benefits of asymmetric augmented paradigm under different GNN architectures, rather than studying individual asymmetric augmented GNN (A2GNN) instance, we then develop an auto-search engine called Asymmetric Augmented Graph Neural Architecture Search (A2GNAS) to save human efforts. We empirically validate our asymmetric augmented paradigm on multiple graph classification benchmarks, and demonstrate that representative A2GNN instances automatically discovered by our A2GNAS method achieve state-of-the-art performance compared with competitive baselines. Our codes are available at: https://github.com/csubigdata-Organization/A2GNAS.
4361.
Wang, Weibo; Li, Hua
NAS FD Lung: A novel lung assist diagnostic system based on neural architecture search Journal Article
In: Biomedical Signal Processing and Control, vol. 100, pp. 107022, 2025, ISSN: 1746-8094.
@article{WANG2025107022,
title = {NAS FD Lung: A novel lung assist diagnostic system based on neural architecture search},
author = {Weibo Wang and Hua Li},
url = {https://www.sciencedirect.com/science/article/pii/S1746809424010802},
doi = {https://doi.org/10.1016/j.bspc.2024.107022},
issn = {1746-8094},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
journal = {Biomedical Signal Processing and Control},
volume = {100},
pages = {107022},
abstract = {In the detection and recognition of lung nodules, pulmonary nodules vary in size and shape and contain many similar tissues and organs around them, leading to the problems of both missed detection and false detection in existing detection algorithms. Designing proprietary detection and recognition networks manually requires substantial professional expertise. This process is time-consuming and labour-intensive and leads to issues like parameter redundancy and improper feature selection. Therefore, this paper proposes a new pulmonary CAD (computer-aided diagnosis) system for pulmonary nodules, NAS FD Lung (Using the NAS approach to search deep FPN and DPN networks), that can automatically learn and generate a deep learning network tailored to pulmonary nodule detection and recognition task requirements. NAS FD Lung aims to use automatic search to generate deep learning networks in the auxiliary diagnosis of pulmonary nodules to replace the manual design of deep learning networks. NAS FD Lung comprises two automatic search networks: BM NAS-FPN (Using NAS methods to search for deep FPN structures with Binary operation and Matrix multiplication fusion methods) network for nodule detection and NAS-A-DPN (Using the NAS approach to search deep DPN networks with attention mechanism) for nodule identification. The proposed technique is tested on the LUNA16 dataset, and the experimental results show that the model is superior to many existing state-of-the-art approaches. The detection accuracy of lung nodules is 98.23%. Regarding the lung nodules classification, the accuracy, specificity, sensitivity, and AUC values achieved 96.32%,97.14%,95.82%, and 98.33%, respectively.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
In the detection and recognition of lung nodules, pulmonary nodules vary in size and shape and contain many similar tissues and organs around them, leading to the problems of both missed detection and false detection in existing detection algorithms. Designing proprietary detection and recognition networks manually requires substantial professional expertise. This process is time-consuming and labour-intensive and leads to issues like parameter redundancy and improper feature selection. Therefore, this paper proposes a new pulmonary CAD (computer-aided diagnosis) system for pulmonary nodules, NAS FD Lung (Using the NAS approach to search deep FPN and DPN networks), that can automatically learn and generate a deep learning network tailored to pulmonary nodule detection and recognition task requirements. NAS FD Lung aims to use automatic search to generate deep learning networks in the auxiliary diagnosis of pulmonary nodules to replace the manual design of deep learning networks. NAS FD Lung comprises two automatic search networks: BM NAS-FPN (Using NAS methods to search for deep FPN structures with Binary operation and Matrix multiplication fusion methods) network for nodule detection and NAS-A-DPN (Using the NAS approach to search deep DPN networks with attention mechanism) for nodule identification. The proposed technique is tested on the LUNA16 dataset, and the experimental results show that the model is superior to many existing state-of-the-art approaches. The detection accuracy of lung nodules is 98.23%. Regarding the lung nodules classification, the accuracy, specificity, sensitivity, and AUC values achieved 96.32%,97.14%,95.82%, and 98.33%, respectively.
4360.
Solis-Martin, David; Galan-Paez, Juan; Borrego-Diaz, Joaquin
Bayesian Model Selection Pruning in Predictive Maintenance Proceedings Article
In: Quintián, Héctor; Corchado, Emilio; Lora, Alicia Troncoso; García, Hilde Pérez; Pérez, Esteban Jove; Rolle, José Luis Calvo; de Pisón, Francisco Javier Martínez; Bringas, Pablo García; Álvarez, Francisco Martínez; Herrero, Álvaro; Fosci, Paolo (Ed.): Hybrid Artificial Intelligent Systems, pp. 263–274, Springer Nature Switzerland, Cham, 2025, ISBN: 978-3-031-74183-8.
@inproceedings{10.1007/978-3-031-74183-8_22,
title = {Bayesian Model Selection Pruning in Predictive Maintenance},
author = {David Solis-Martin and Juan Galan-Paez and Joaquin Borrego-Diaz},
editor = {Héctor Quintián and Emilio Corchado and Alicia Troncoso Lora and Hilde Pérez García and Esteban Jove Pérez and José Luis Calvo Rolle and Francisco Javier Martínez de Pisón and Pablo García Bringas and Francisco Martínez Álvarez and Álvaro Herrero and Paolo Fosci},
url = {https://link.springer.com/chapter/10.1007/978-3-031-74183-8_22},
isbn = {978-3-031-74183-8},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
booktitle = {Hybrid Artificial Intelligent Systems},
pages = {263–274},
publisher = {Springer Nature Switzerland},
address = {Cham},
abstract = {Deep Neural Network architecture design significantly impacts the final model performance. The process of searching for optimal architectures, known as Neural Architecture Search (NAS), involves training and evaluating an important number of models. Therefore, mechanisms to reduce the resources required for NAS are highly valuable.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Deep Neural Network architecture design significantly impacts the final model performance. The process of searching for optimal architectures, known as Neural Architecture Search (NAS), involves training and evaluating an important number of models. Therefore, mechanisms to reduce the resources required for NAS are highly valuable.
4359.
Solis-Martin, David; Galan-Paez, Juan; Borrego-Diaz, Joaquin
Bayesian Model Selection Pruning in Predictive Maintenance Proceedings Article
In: Quintián, Héctor; Corchado, Emilio; Lora, Alicia Troncoso; García, Hilde Pérez; Pérez, Esteban Jove; Rolle, José Luis Calvo; de Pisón, Francisco Javier Martínez; Bringas, Pablo García; Álvarez, Francisco Martínez; Herrero, Álvaro; Fosci, Paolo (Ed.): Hybrid Artificial Intelligent Systems, pp. 263–274, Springer Nature Switzerland, Cham, 2025, ISBN: 978-3-031-74183-8.
@inproceedings{10.1007/978-3-031-74183-8_22b,
title = {Bayesian Model Selection Pruning in Predictive Maintenance},
author = {David Solis-Martin and Juan Galan-Paez and Joaquin Borrego-Diaz},
editor = {Héctor Quintián and Emilio Corchado and Alicia Troncoso Lora and Hilde Pérez García and Esteban Jove Pérez and José Luis Calvo Rolle and Francisco Javier Martínez de Pisón and Pablo García Bringas and Francisco Martínez Álvarez and Álvaro Herrero and Paolo Fosci},
url = {https://link.springer.com/chapter/10.1007/978-3-031-74183-8_22},
isbn = {978-3-031-74183-8},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
booktitle = {Hybrid Artificial Intelligent Systems},
pages = {263–274},
publisher = {Springer Nature Switzerland},
address = {Cham},
abstract = {Deep Neural Network architecture design significantly impacts the final model performance. The process of searching for optimal architectures, known as Neural Architecture Search (NAS), involves training and evaluating an important number of models. Therefore, mechanisms to reduce the resources required for NAS are highly valuable.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Deep Neural Network architecture design significantly impacts the final model performance. The process of searching for optimal architectures, known as Neural Architecture Search (NAS), involves training and evaluating an important number of models. Therefore, mechanisms to reduce the resources required for NAS are highly valuable.
4358.
Garcia-Garcia, Cosijopii; Derbel, Bilel; Morales-Reyes, Alicia; Escalante, Hugo Jair
Speeding up the Multi-objective NAS Through Incremental Learning Proceedings Article
In: Martínez-Villaseñor, Lourdes; Ochoa-Ruiz, Gilberto (Ed.): Advances in Soft Computing, pp. 3–15, Springer Nature Switzerland, Cham, 2025, ISBN: 978-3-031-75543-9.
@inproceedings{10.1007/978-3-031-75543-9_1,
title = {Speeding up the Multi-objective NAS Through Incremental Learning},
author = {Cosijopii Garcia-Garcia and Bilel Derbel and Alicia Morales-Reyes and Hugo Jair Escalante},
editor = {Lourdes Martínez-Villaseñor and Gilberto Ochoa-Ruiz},
url = {https://link.springer.com/chapter/10.1007/978-3-031-75543-9_1},
isbn = {978-3-031-75543-9},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
booktitle = {Advances in Soft Computing},
pages = {3–15},
publisher = {Springer Nature Switzerland},
address = {Cham},
abstract = {Deep neural networks (DNNs), particularly convolutional neural networks (CNNs), have garnered significant attention in recent years for addressing a wide range of challenges in image processing and computer vision. Neural architecture search (NAS) has emerged as a crucial field aiming to automate the design and configuration of CNN models. In this paper, we propose a novel strategy to speed up the performance estimation of neural architectures by gradually increasing the size of the training set used for evaluation as the search progresses. We evaluate this approach using the CGP-NASV2 model, a multi-objective NAS method, on the CIFAR-100 dataset. Experimental results demonstrate a notable acceleration in the search process, achieving a speedup of 4.6 times compared to the baseline. Despite using limited data in the early stages, our proposed method effectively guides the search towards competitive architectures. This study highlights the efficacy of leveraging lower-fidelity estimates in NAS and paves the way for further research into accelerating the design of efficient CNN architectures.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Deep neural networks (DNNs), particularly convolutional neural networks (CNNs), have garnered significant attention in recent years for addressing a wide range of challenges in image processing and computer vision. Neural architecture search (NAS) has emerged as a crucial field aiming to automate the design and configuration of CNN models. In this paper, we propose a novel strategy to speed up the performance estimation of neural architectures by gradually increasing the size of the training set used for evaluation as the search progresses. We evaluate this approach using the CGP-NASV2 model, a multi-objective NAS method, on the CIFAR-100 dataset. Experimental results demonstrate a notable acceleration in the search process, achieving a speedup of 4.6 times compared to the baseline. Despite using limited data in the early stages, our proposed method effectively guides the search towards competitive architectures. This study highlights the efficacy of leveraging lower-fidelity estimates in NAS and paves the way for further research into accelerating the design of efficient CNN architectures.
4357.
Liu, Wenbo; Deng, Tao; An, Rui; Yan, Fei
DARTS-CGW: Research on Differentiable Neural Architecture Search Algorithm Based on Coarse Gradient Weighting Proceedings Article
In: Lin, Zhouchen; Cheng, Ming-Ming; He, Ran; Ubul, Kurban; Silamu, Wushouer; Zha, Hongbin; Zhou, Jie; Liu, Cheng-Lin (Ed.): Pattern Recognition and Computer Vision, pp. 31–44, Springer Nature Singapore, Singapore, 2025, ISBN: 978-981-97-8502-5.
@inproceedings{10.1007/978-981-97-8502-5_3,
title = {DARTS-CGW: Research on Differentiable Neural Architecture Search Algorithm Based on Coarse Gradient Weighting},
author = {Wenbo Liu and Tao Deng and Rui An and Fei Yan},
editor = {Zhouchen Lin and Ming-Ming Cheng and Ran He and Kurban Ubul and Wushouer Silamu and Hongbin Zha and Jie Zhou and Cheng-Lin Liu},
url = {https://link.springer.com/chapter/10.1007/978-981-97-8502-5_3},
isbn = {978-981-97-8502-5},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
booktitle = {Pattern Recognition and Computer Vision},
pages = {31–44},
publisher = {Springer Nature Singapore},
address = {Singapore},
abstract = {Differential architecture search (DARTS) has emerged as a prominent research area, yet it grapples with a longstanding challenge: the discretization discrepancy problem. This issue directly impedes the search for an optimal model architecture and undermines search algorithm performance. To alleviate this issue, we propose a novel coarse gradient weighting algorithm. Our proposed algorithm has the capability to simulate the discretization process, wherein the architectural parameters move toward both ends. And we integrate this discretization process into the training phase of the architectural parameters, enabling the model to adapt to the discretization process in a trial-and-error fashion. Specifically, based on the architectural parameters in training, we divide the candidate operations into two regions, i.e., the easy-to-select region and the hard-to-be-selected region. The different weighting strategies are implemented in different regions so that the architectural parameters are pushed to the ends. The processed architecture parameters are used for training, which is equivalent to introducing the discretization process into the search phase. Additionally, we use the coarse gradient algorithm to optimize the updating process of the weighting algorithm and theoretically justify the rationality of the coarse gradient weighting algorithm. Extensive experimental results demonstrate that our proposed method can improve the performance of the searched model and make DARTS more robust without adding additional search time.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Differential architecture search (DARTS) has emerged as a prominent research area, yet it grapples with a longstanding challenge: the discretization discrepancy problem. This issue directly impedes the search for an optimal model architecture and undermines search algorithm performance. To alleviate this issue, we propose a novel coarse gradient weighting algorithm. Our proposed algorithm has the capability to simulate the discretization process, wherein the architectural parameters move toward both ends. And we integrate this discretization process into the training phase of the architectural parameters, enabling the model to adapt to the discretization process in a trial-and-error fashion. Specifically, based on the architectural parameters in training, we divide the candidate operations into two regions, i.e., the easy-to-select region and the hard-to-be-selected region. The different weighting strategies are implemented in different regions so that the architectural parameters are pushed to the ends. The processed architecture parameters are used for training, which is equivalent to introducing the discretization process into the search phase. Additionally, we use the coarse gradient algorithm to optimize the updating process of the weighting algorithm and theoretically justify the rationality of the coarse gradient weighting algorithm. Extensive experimental results demonstrate that our proposed method can improve the performance of the searched model and make DARTS more robust without adding additional search time.
4356.
Man, Wenxing; Xu, Liming; He, Chunlin
Evolutionary architecture search for generative adversarial networks using an aging mechanism-based strategy Journal Article
In: Neural Networks, vol. 181, pp. 106877, 2025, ISSN: 0893-6080.
@article{MAN2025106877,
title = {Evolutionary architecture search for generative adversarial networks using an aging mechanism-based strategy},
author = {Wenxing Man and Liming Xu and Chunlin He},
url = {https://www.sciencedirect.com/science/article/pii/S0893608024008050},
doi = {https://doi.org/10.1016/j.neunet.2024.106877},
issn = {0893-6080},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
journal = {Neural Networks},
volume = {181},
pages = {106877},
abstract = {Generative Adversarial Networks (GANs) have emerged as a key technology in artificial intelligence, especially in image generation. However, traditionally hand-designed GAN architectures often face significant training stability challenges, which are effectively addressed by our Evolutionary Neural Architecture Search (ENAS) algorithm for GANs, named EAMGAN. This one-shot model automates the design of GAN architectures and employs an Operation Importance Metric (OIM) to enhance training stability. It also incorporates an aging mechanism to optimize the selection process during architecture search. Additionally, the use of a non-dominated sorting algorithm ensures the generation of Pareto-optimal solutions, promoting diversity and preventing premature convergence. We evaluated our method on benchmark datasets, and the results demonstrate that EAMGAN is highly competitive in terms of efficiency and performance. Our method identified an architecture achieving Inception Scores (IS) of 8.83±0.13 and Fréchet Inception Distance (FID) of 9.55 on CIFAR-10 with only 0.66 GPU days. Results on the STL-10, CIFAR-100, and ImageNet32 datasets further demonstrate the robust portability of our architecture.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Generative Adversarial Networks (GANs) have emerged as a key technology in artificial intelligence, especially in image generation. However, traditionally hand-designed GAN architectures often face significant training stability challenges, which are effectively addressed by our Evolutionary Neural Architecture Search (ENAS) algorithm for GANs, named EAMGAN. This one-shot model automates the design of GAN architectures and employs an Operation Importance Metric (OIM) to enhance training stability. It also incorporates an aging mechanism to optimize the selection process during architecture search. Additionally, the use of a non-dominated sorting algorithm ensures the generation of Pareto-optimal solutions, promoting diversity and preventing premature convergence. We evaluated our method on benchmark datasets, and the results demonstrate that EAMGAN is highly competitive in terms of efficiency and performance. Our method identified an architecture achieving Inception Scores (IS) of 8.83±0.13 and Fréchet Inception Distance (FID) of 9.55 on CIFAR-10 with only 0.66 GPU days. Results on the STL-10, CIFAR-100, and ImageNet32 datasets further demonstrate the robust portability of our architecture.
4355.
Zhong, Rui; Xu, Yuefeng; Zhang, Chao; Yu, Jun
Efficient multiplayer battle game optimizer for numerical optimization and adversarial robust neural architecture search Journal Article
In: Alexandria Engineering Journal, vol. 113, pp. 150-168, 2025, ISSN: 1110-0168.
@article{ZHONG2025150,
title = {Efficient multiplayer battle game optimizer for numerical optimization and adversarial robust neural architecture search},
author = {Rui Zhong and Yuefeng Xu and Chao Zhang and Jun Yu},
url = {https://www.sciencedirect.com/science/article/pii/S1110016824014935},
doi = {https://doi.org/10.1016/j.aej.2024.11.035},
issn = {1110-0168},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
journal = {Alexandria Engineering Journal},
volume = {113},
pages = {150-168},
abstract = {This paper introduces a novel metaheuristic algorithm, known as the efficient multiplayer battle game optimizer (EMBGO), specifically designed for addressing complex numerical optimization tasks. The motivation behind this research stems from the need to rectify identified shortcomings in the original MBGO, particularly in search operators during the movement phase, as revealed through ablation experiments. EMBGO mitigates these limitations by integrating the movement and battle phases to simplify the original optimization framework and improve search efficiency. Besides, two efficient search operators: differential mutation and Lévy flight are introduced to increase the diversity of the population. To evaluate the performance of EMBGO comprehensively and fairly, numerical experiments are conducted on benchmark functions such as CEC2017, CEC2020, and CEC2022, as well as engineering problems. Twelve well-established MA approaches serve as competitor algorithms for comparison. Furthermore, we apply the proposed EMBGO to the complex adversarial robust neural architecture search (ARNAS) tasks and explore its robustness and scalability. The experimental results and statistical analyses confirm the efficiency and effectiveness of EMBGO across various optimization tasks. As a potential optimization technique, EMBGO holds promise for diverse applications in real-world problems and deep learning scenarios. The source code of EMBGO is made available in https://github.com/RuiZhong961230/EMBGO.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
This paper introduces a novel metaheuristic algorithm, known as the efficient multiplayer battle game optimizer (EMBGO), specifically designed for addressing complex numerical optimization tasks. The motivation behind this research stems from the need to rectify identified shortcomings in the original MBGO, particularly in search operators during the movement phase, as revealed through ablation experiments. EMBGO mitigates these limitations by integrating the movement and battle phases to simplify the original optimization framework and improve search efficiency. Besides, two efficient search operators: differential mutation and Lévy flight are introduced to increase the diversity of the population. To evaluate the performance of EMBGO comprehensively and fairly, numerical experiments are conducted on benchmark functions such as CEC2017, CEC2020, and CEC2022, as well as engineering problems. Twelve well-established MA approaches serve as competitor algorithms for comparison. Furthermore, we apply the proposed EMBGO to the complex adversarial robust neural architecture search (ARNAS) tasks and explore its robustness and scalability. The experimental results and statistical analyses confirm the efficiency and effectiveness of EMBGO across various optimization tasks. As a potential optimization technique, EMBGO holds promise for diverse applications in real-world problems and deep learning scenarios. The source code of EMBGO is made available in https://github.com/RuiZhong961230/EMBGO.
4354.
Wang, Yu; Fan, Jiahao; Sun, Yanan
Classification of sewer pipe defects based on an automatically designed convolutional neural network Journal Article
In: Expert Systems with Applications, vol. 264, pp. 125806, 2025, ISSN: 0957-4174.
@article{WANG2025125806,
title = {Classification of sewer pipe defects based on an automatically designed convolutional neural network},
author = {Yu Wang and Jiahao Fan and Yanan Sun},
url = {https://www.sciencedirect.com/science/article/pii/S0957417424026733},
doi = {https://doi.org/10.1016/j.eswa.2024.125806},
issn = {0957-4174},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
journal = {Expert Systems with Applications},
volume = {264},
pages = {125806},
abstract = {Accurate classification of sewer pipe defects allows for timely prevention of system failures, preventing environmental pollution caused by sewage leakage, thereby safeguarding public health and ensuring urban infrastructure operates normally. However, existing methods with deep convolutional neural networks not only require a significant amount of time to design network architectures, but also demand substantial computational resources. Therefore, this paper automatically designed high-performance and lightweight AutoSewerNet based on neural architecture search (NAS) for classification of sewer pipe defects. First, Super-net is designed to enrich the diversity of search space and include high-performance network architecture. Second, to reduce the search time of NAS, a gradient-based search strategy is designed. Third, AutoSewerNet is designed to be lightweight and perform real-time inspections of sewer pipes. Fourth, weight balance is introduced to solve the imbalanced dataset problem. Experimental results demonstrate that AutoSewerNet achieved an F1-score of 0.6251 on the benchmark dataset, which is far superior to that of ResNet-50 (F1-score: 0.4523) and InceptionV3 (F1-score: 0.4611). Moreover, AutoSewerNet requires only 11.6% of VGG-16. Thus, AutoSewerNet is better than state-of-the-art methods. Our source code and models are anonymously available at https://anonymous.4open.science/r/ASN-26AB/.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Accurate classification of sewer pipe defects allows for timely prevention of system failures, preventing environmental pollution caused by sewage leakage, thereby safeguarding public health and ensuring urban infrastructure operates normally. However, existing methods with deep convolutional neural networks not only require a significant amount of time to design network architectures, but also demand substantial computational resources. Therefore, this paper automatically designed high-performance and lightweight AutoSewerNet based on neural architecture search (NAS) for classification of sewer pipe defects. First, Super-net is designed to enrich the diversity of search space and include high-performance network architecture. Second, to reduce the search time of NAS, a gradient-based search strategy is designed. Third, AutoSewerNet is designed to be lightweight and perform real-time inspections of sewer pipes. Fourth, weight balance is introduced to solve the imbalanced dataset problem. Experimental results demonstrate that AutoSewerNet achieved an F1-score of 0.6251 on the benchmark dataset, which is far superior to that of ResNet-50 (F1-score: 0.4523) and InceptionV3 (F1-score: 0.4611). Moreover, AutoSewerNet requires only 11.6% of VGG-16. Thus, AutoSewerNet is better than state-of-the-art methods. Our source code and models are anonymously available at https://anonymous.4open.science/r/ASN-26AB/.
4353.
Fu, Wei; Lou, Wenqi; Qin, Yunji; Gong, Lei; Wang, Chao; Zhou, Xuehai
MFNAS: Multi-fidelity Exploration in Neural Architecture Search with Stable Zero-Shot Proxy Proceedings Article
In: Hadfi, Rafik; Anthony, Patricia; Sharma, Alok; Ito, Takayuki; Bai, Quan (Ed.): PRICAI 2024: Trends in Artificial Intelligence, pp. 348–360, Springer Nature Singapore, Singapore, 2025, ISBN: 978-981-96-0116-5.
@inproceedings{10.1007/978-981-96-0116-5_29,
title = {MFNAS: Multi-fidelity Exploration in Neural Architecture Search with Stable Zero-Shot Proxy},
author = {Wei Fu and Wenqi Lou and Yunji Qin and Lei Gong and Chao Wang and Xuehai Zhou},
editor = {Rafik Hadfi and Patricia Anthony and Alok Sharma and Takayuki Ito and Quan Bai},
isbn = {978-981-96-0116-5},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
booktitle = {PRICAI 2024: Trends in Artificial Intelligence},
pages = {348–360},
publisher = {Springer Nature Singapore},
address = {Singapore},
abstract = {Neural architecture search (NAS) automates the design of neural networks for specific tasks. Recently, zero-shot NAS has attracted much attention. Unlike traditional NAS, which relies on training to rank architectures, zero-shot NAS uses gradients or activation information to evaluate architecture performance. However, existing zero-shot NAS methods are limited by their inconsistent architecture ranking and the evaluation bias of their search algorithm, making it challenging to discover networks with high accuracy efficiently. To address this dilemma, this paper proposes an efficient and stable search framework for zero-shot NAS. Firstly, we design a stable zero-shot proxy, which achieves good consistency with network accuracy by utilizing filtered gradient information. On this basis, we employ a multi-fidelity evolutionary algorithm for efficient exploration. This algorithm utilizes multi-fidelity proxies to correct the bias towards certain types of networks and enhances the ability to distinguish high-performing architectures, achieving rapid convergence through performance-directed multi-point crossover and mutation. Experimental results conducted on NATS-Bench demonstrate that our framework can discover high-performance architectures within minutes of GPU time, outperforming existing training-free and training-based NAS methods. The code is available at https://github.com/mine7777/MFNAS.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Neural architecture search (NAS) automates the design of neural networks for specific tasks. Recently, zero-shot NAS has attracted much attention. Unlike traditional NAS, which relies on training to rank architectures, zero-shot NAS uses gradients or activation information to evaluate architecture performance. However, existing zero-shot NAS methods are limited by their inconsistent architecture ranking and the evaluation bias of their search algorithm, making it challenging to discover networks with high accuracy efficiently. To address this dilemma, this paper proposes an efficient and stable search framework for zero-shot NAS. Firstly, we design a stable zero-shot proxy, which achieves good consistency with network accuracy by utilizing filtered gradient information. On this basis, we employ a multi-fidelity evolutionary algorithm for efficient exploration. This algorithm utilizes multi-fidelity proxies to correct the bias towards certain types of networks and enhances the ability to distinguish high-performing architectures, achieving rapid convergence through performance-directed multi-point crossover and mutation. Experimental results conducted on NATS-Bench demonstrate that our framework can discover high-performance architectures within minutes of GPU time, outperforming existing training-free and training-based NAS methods. The code is available at https://github.com/mine7777/MFNAS.
4352.
Verma, Sahil; Kumar, Prabhat; Singh, Jyoti Prakash
MLP-GNAS: Meta-learning-based predictor-assisted Genetic Neural Architecture Search system Journal Article
In: Applied Soft Computing, vol. 169, pp. 112527, 2025, ISSN: 1568-4946.
@article{VERMA2025112527,
title = {MLP-GNAS: Meta-learning-based predictor-assisted Genetic Neural Architecture Search system},
author = {Sahil Verma and Prabhat Kumar and Jyoti Prakash Singh},
url = {https://www.sciencedirect.com/science/article/pii/S1568494624013012},
doi = {https://doi.org/10.1016/j.asoc.2024.112527},
issn = {1568-4946},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
journal = {Applied Soft Computing},
volume = {169},
pages = {112527},
abstract = {Convolutional neural networks achieve state-of-the-art results on many image classification tasks. However, according to the No Free Lunch Theorem, no single model performs optimally for all the datasets. Hence, the manual development of these models using the hit-and-trial approach requires high computational overhead. Transfer learning mitigates this issue to some extent by transferring knowledge learned from similar domain tasks to target tasks through model architecture and weights. Yet, the model remains designed for the source task, and using the same model on the target task may result in over-parameterization or sub-optimal results. Henceforth, the current work proposes a meta-learning-based predictor-assisted genetic Neural Architecture Search (MLP-GNAS) system to automate the model generation process for plant disease detection tasks to validate its efficacy in real-world scenarios. To this cause, the MLP-GNAS system employs a meta-learning component to recommend the top 3 suitable models for a target dataset, followed by a genetic algorithm-based NAS to fine-tune the recommended model. In addition, the proposed approach uses a CNN-based performance predictor designed to discard models unlikely to surpass the optimal performance recorded thus far. Further, extensive comparative experiments demonstrate that MLP-GNAS-generated models outperform 14 state-of-the-art models on two distinct plant disease datasets with an accuracy of 96.94% and 92.04%. Moreover, further comparison with the manually fine-tuned and NAS-generated models reveals that the proposed system outperforms both to obtain an accuracy of 98.35% and 99.95% on respective datasets.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Convolutional neural networks achieve state-of-the-art results on many image classification tasks. However, according to the No Free Lunch Theorem, no single model performs optimally for all the datasets. Hence, the manual development of these models using the hit-and-trial approach requires high computational overhead. Transfer learning mitigates this issue to some extent by transferring knowledge learned from similar domain tasks to target tasks through model architecture and weights. Yet, the model remains designed for the source task, and using the same model on the target task may result in over-parameterization or sub-optimal results. Henceforth, the current work proposes a meta-learning-based predictor-assisted genetic Neural Architecture Search (MLP-GNAS) system to automate the model generation process for plant disease detection tasks to validate its efficacy in real-world scenarios. To this cause, the MLP-GNAS system employs a meta-learning component to recommend the top 3 suitable models for a target dataset, followed by a genetic algorithm-based NAS to fine-tune the recommended model. In addition, the proposed approach uses a CNN-based performance predictor designed to discard models unlikely to surpass the optimal performance recorded thus far. Further, extensive comparative experiments demonstrate that MLP-GNAS-generated models outperform 14 state-of-the-art models on two distinct plant disease datasets with an accuracy of 96.94% and 92.04%. Moreover, further comparison with the manually fine-tuned and NAS-generated models reveals that the proposed system outperforms both to obtain an accuracy of 98.35% and 99.95% on respective datasets.
4351.
Oloulade, Babatounde Moctard; Gao, Jianliang; Chen, Jiamin; Al-Sabri, Raeed; Wu, Zhenpeng; Abdullah, Monir
Shapley-guided pruning for efficient graph neural architecture prediction in distributed learning environments Journal Article
In: Information Sciences, vol. 695, pp. 121695, 2025, ISSN: 0020-0255.
@article{OLOULADE2025121695,
title = {Shapley-guided pruning for efficient graph neural architecture prediction in distributed learning environments},
author = {Babatounde Moctard Oloulade and Jianliang Gao and Jiamin Chen and Raeed Al-Sabri and Zhenpeng Wu and Monir Abdullah},
url = {https://www.sciencedirect.com/science/article/pii/S0020025524016098},
doi = {https://doi.org/10.1016/j.ins.2024.121695},
issn = {0020-0255},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
journal = {Information Sciences},
volume = {695},
pages = {121695},
abstract = {Graph neural architecture search (GNAS) methods have made significant strides in predictive modeling; however, challenges persist in optimizing these approaches for improved accuracy and scalability. In this paper, we propose GraphNAP++, an innovative framework that combines Shapley-value-guided search space pruning with data-centric distributed learning (DCDL) to address the scalability and efficiency limitations of existing GNAS methodologies. The proposed method begins by selecting a subset of architectures from the search space, which are evaluated on a graph validation dataset using DCDL. These architectures are encoded, and a neural predictor is trained to predict their performance. Shapley values are utilized to prune the search space by retaining only the most influential architectures. The neural predictor then estimates the performance of all architectures within the reduced search space, and the highest-performing architectures are selected for final evaluation. Experimental results on benchmark datasets in distributed learning environments demonstrate the effectiveness of GraphNAP++ in both graph and node classification tasks, highlighting its potential to advance the state-of-the-art for graph neural networks.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Graph neural architecture search (GNAS) methods have made significant strides in predictive modeling; however, challenges persist in optimizing these approaches for improved accuracy and scalability. In this paper, we propose GraphNAP++, an innovative framework that combines Shapley-value-guided search space pruning with data-centric distributed learning (DCDL) to address the scalability and efficiency limitations of existing GNAS methodologies. The proposed method begins by selecting a subset of architectures from the search space, which are evaluated on a graph validation dataset using DCDL. These architectures are encoded, and a neural predictor is trained to predict their performance. Shapley values are utilized to prune the search space by retaining only the most influential architectures. The neural predictor then estimates the performance of all architectures within the reduced search space, and the highest-performing architectures are selected for final evaluation. Experimental results on benchmark datasets in distributed learning environments demonstrate the effectiveness of GraphNAP++ in both graph and node classification tasks, highlighting its potential to advance the state-of-the-art for graph neural networks.
4350.
Chen, Jiakun; Xu, Jie; Hu, Jiahui; Qiao, Liqiang; Wang, Shuo; Huang, Feiran; Li, Chaozhuo
Context-Aware Structural Adaptive Graph Neural Networks Proceedings Article
In: Hadfi, Rafik; Anthony, Patricia; Sharma, Alok; Ito, Takayuki; Bai, Quan (Ed.): PRICAI 2024: Trends in Artificial Intelligence, pp. 467–479, Springer Nature Singapore, Singapore, 2025, ISBN: 978-981-96-0116-5.
@inproceedings{10.1007/978-981-96-0116-5_39,
title = {Context-Aware Structural Adaptive Graph Neural Networks},
author = {Jiakun Chen and Jie Xu and Jiahui Hu and Liqiang Qiao and Shuo Wang and Feiran Huang and Chaozhuo Li},
editor = {Rafik Hadfi and Patricia Anthony and Alok Sharma and Takayuki Ito and Quan Bai},
url = {https://link.springer.com/chapter/10.1007/978-981-96-0116-5_39},
isbn = {978-981-96-0116-5},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
booktitle = {PRICAI 2024: Trends in Artificial Intelligence},
pages = {467–479},
publisher = {Springer Nature Singapore},
address = {Singapore},
abstract = {Graph-based data structures are prevalent in various real-world applications, for example, protein molecules and social connection networks, necessitating effective representation learning techniques. Graph Neural Networks (GNNs) have demonstrated significant advancements in tasks like node classification and social network analysis through recursive information aggregation. However, current GNN approaches are predominantly static, lacking adaptability to specific graph structures. Inspired by Neural Architecture Search (NAS) in designing dataset-specific architectures, we propose Context-Aware Structure Adaptive Graph Neural Networks (CAS-GNN). This framework is capable of automatically selecting the appropriate aggregator for each node which is determined by both node attributes and local contextual information. The selection is formulated as the Markov Decision Process (MDP) optimized via Deep-Q-Network (DQN) training. Our contributions include a flexible framework incorporating various aggregators for individual nodes based on their attributes and local context, improved performance through node-specific aggregator selection, and extensive experimental validation demonstrating the effectiveness of CAS-GNN on multiple real-world datasets.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Graph-based data structures are prevalent in various real-world applications, for example, protein molecules and social connection networks, necessitating effective representation learning techniques. Graph Neural Networks (GNNs) have demonstrated significant advancements in tasks like node classification and social network analysis through recursive information aggregation. However, current GNN approaches are predominantly static, lacking adaptability to specific graph structures. Inspired by Neural Architecture Search (NAS) in designing dataset-specific architectures, we propose Context-Aware Structure Adaptive Graph Neural Networks (CAS-GNN). This framework is capable of automatically selecting the appropriate aggregator for each node which is determined by both node attributes and local contextual information. The selection is formulated as the Markov Decision Process (MDP) optimized via Deep-Q-Network (DQN) training. Our contributions include a flexible framework incorporating various aggregators for individual nodes based on their attributes and local context, improved performance through node-specific aggregator selection, and extensive experimental validation demonstrating the effectiveness of CAS-GNN on multiple real-world datasets.
4349.
Reddy, P. N. Aravinda; Ramachandra, Raghavendra; Rao, K. Sreenivasa; Mitra, Pabitra
NeuralMultiling: A Novel Neural Architecture Search for Smartphone Based Multilingual Speaker Verification Proceedings Article
In: Antonacopoulos, Apostolos; Chaudhuri, Subhasis; Chellappa, Rama; Liu, Cheng-Lin; Bhattacharya, Saumik; Pal, Umapada (Ed.): Pattern Recognition, pp. 406–423, Springer Nature Switzerland, Cham, 2025, ISBN: 978-3-031-78341-8.
@inproceedings{10.1007/978-3-031-78341-8_26,
title = {NeuralMultiling: A Novel Neural Architecture Search for Smartphone Based Multilingual Speaker Verification},
author = {P. N. Aravinda Reddy and Raghavendra Ramachandra and K. Sreenivasa Rao and Pabitra Mitra},
editor = {Apostolos Antonacopoulos and Subhasis Chaudhuri and Rama Chellappa and Cheng-Lin Liu and Saumik Bhattacharya and Umapada Pal},
url = {https://link.springer.com/chapter/10.1007/978-3-031-78341-8_26},
isbn = {978-3-031-78341-8},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
booktitle = {Pattern Recognition},
pages = {406–423},
publisher = {Springer Nature Switzerland},
address = {Cham},
abstract = {Multilingual speaker verification introduces the challenge of verifying a speaker in multiple languages. Existing systems were built using i-vector/x-vector approaches along with Bi-LSTMs, which were trained to discriminate speakers, irrespective of the language. Instead of exploring the design space manually, we propose a neural architecture search for multilingual speaker verification suitable for mobile devices, called NeuralMultiling. First, our algorithm searches for an optimal operational combination of neural cells with different architectures for normal cells and reduction cells and then derives a CNN model by stacking neural cells. Using the derived architecture, we performed two different studies:1) language agnostic condition and 2) interoperability between languages and devices on the publicly available Multilingual Audio-Visual Smartphone (MAVS) dataset. The experimental results suggest that the derived architecture significantly outperforms the existing Autospeech method by a 5–6% reduction in the Equal Error Rate (EER) with fewer model parameters.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Multilingual speaker verification introduces the challenge of verifying a speaker in multiple languages. Existing systems were built using i-vector/x-vector approaches along with Bi-LSTMs, which were trained to discriminate speakers, irrespective of the language. Instead of exploring the design space manually, we propose a neural architecture search for multilingual speaker verification suitable for mobile devices, called NeuralMultiling. First, our algorithm searches for an optimal operational combination of neural cells with different architectures for normal cells and reduction cells and then derives a CNN model by stacking neural cells. Using the derived architecture, we performed two different studies:1) language agnostic condition and 2) interoperability between languages and devices on the publicly available Multilingual Audio-Visual Smartphone (MAVS) dataset. The experimental results suggest that the derived architecture significantly outperforms the existing Autospeech method by a 5–6% reduction in the Equal Error Rate (EER) with fewer model parameters.
4348.
Li, Jie; Wang, Yuxia; Wang, Yifan; Yu, Ruiyun; Wang, Xingwei
HN-Darts:Hybrid Network Differentiable Architecture Search for Industrial Scenarios Proceedings Article
In: Hadfi, Rafik; Anthony, Patricia; Sharma, Alok; Ito, Takayuki; Bai, Quan (Ed.): PRICAI 2024: Trends in Artificial Intelligence, pp. 322–327, Springer Nature Singapore, Singapore, 2025, ISBN: 978-981-96-0116-5.
@inproceedings{10.1007/978-981-96-0116-5_26,
title = {HN-Darts:Hybrid Network Differentiable Architecture Search for Industrial Scenarios},
author = {Jie Li and Yuxia Wang and Yifan Wang and Ruiyun Yu and Xingwei Wang},
editor = {Rafik Hadfi and Patricia Anthony and Alok Sharma and Takayuki Ito and Quan Bai},
url = {https://link.springer.com/chapter/10.1007/978-981-96-0116-5_26},
isbn = {978-981-96-0116-5},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
booktitle = {PRICAI 2024: Trends in Artificial Intelligence},
pages = {322–327},
publisher = {Springer Nature Singapore},
address = {Singapore},
abstract = {Neural architecture search is a powerful tool in image processing, automating model construction and reducing human involvement. However, its deployment on edge devices with limited computing resources is often impeded by the size of large models, a concern overlooked by most NAS methods focused solely on accuracy. We propose a hybrid network search approach that integrates the glore_unit, a novel component that replaces traditional cells to optimize model size without sacrificing accuracy. By leveraging the Differentiable Architecture Search (Darts) and a Googlenet-like hypernet, we've redefined the search space to prioritize compactness and precision, enhanced by a temperature factor for more reliable search selections. Our experiments on cifar10 and ImageNet showcase a model with a 2.35% error rate and 2.76M parameters on cifar10, and a Top-1 error rate of 23.75%, Top-5 error rate of 7.13% with 3.9M parameters on ImageNet, demonstrating SOTA accuracy with a significant reduction in model parameters, making it suitable for environments with constrained computational resources.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Neural architecture search is a powerful tool in image processing, automating model construction and reducing human involvement. However, its deployment on edge devices with limited computing resources is often impeded by the size of large models, a concern overlooked by most NAS methods focused solely on accuracy. We propose a hybrid network search approach that integrates the glore_unit, a novel component that replaces traditional cells to optimize model size without sacrificing accuracy. By leveraging the Differentiable Architecture Search (Darts) and a Googlenet-like hypernet, we've redefined the search space to prioritize compactness and precision, enhanced by a temperature factor for more reliable search selections. Our experiments on cifar10 and ImageNet showcase a model with a 2.35% error rate and 2.76M parameters on cifar10, and a Top-1 error rate of 23.75%, Top-5 error rate of 7.13% with 3.9M parameters on ImageNet, demonstrating SOTA accuracy with a significant reduction in model parameters, making it suitable for environments with constrained computational resources.
4347.
Jin, Xiao; Yu, Wen; Chen, Dai-Wei; Shi, Wei
DFD-NAS: General deepfake detection via efficient neural architecture search Journal Article
In: Neurocomputing, vol. 619, pp. 129129, 2025, ISSN: 0925-2312.
@article{JIN2025129129,
title = {DFD-NAS: General deepfake detection via efficient neural architecture search},
author = {Xiao Jin and Wen Yu and Dai-Wei Chen and Wei Shi},
url = {https://www.sciencedirect.com/science/article/pii/S0925231224019003},
doi = {https://doi.org/10.1016/j.neucom.2024.129129},
issn = {0925-2312},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
journal = {Neurocomputing},
volume = {619},
pages = {129129},
abstract = {As the saying goes, “seeing is believing”. However, with the development of digital face editing tools, we can no longer trust what we can see. Though face forgery detection has made promising progress, most current methods are manually designed by human experts, which is labor-intensive. In this paper, we develop an end-to-end framework based on neural architecture search (NAS) for deepfake detection, which can automatically design network architectures without human intervention. First, a forgery-oriented search space is created to choose appropriate operations for this task, which facilitates the search process in finding the most valuable gradient information for face forgery detection. Second, inspired by the fact that the gap between training error and test error is a good indicator of generalization ability for a classification task, we propose a novel performance estimation metric. This metric encourages the error in the performance estimation phase to be close to the error in the search phase, which guides the search process to select more general models. The cross-dataset search is also considered to develop more general architectures. Eventually, we design a cell cascaded pyramid network (C2PN) for final detection, which aggregates multiscale features for performance improvements. Compared with state-of-the-art networks artificially designed, our method achieves competitive performance in both in-dataset and cross-dataset scenarios.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
As the saying goes, “seeing is believing”. However, with the development of digital face editing tools, we can no longer trust what we can see. Though face forgery detection has made promising progress, most current methods are manually designed by human experts, which is labor-intensive. In this paper, we develop an end-to-end framework based on neural architecture search (NAS) for deepfake detection, which can automatically design network architectures without human intervention. First, a forgery-oriented search space is created to choose appropriate operations for this task, which facilitates the search process in finding the most valuable gradient information for face forgery detection. Second, inspired by the fact that the gap between training error and test error is a good indicator of generalization ability for a classification task, we propose a novel performance estimation metric. This metric encourages the error in the performance estimation phase to be close to the error in the search phase, which guides the search process to select more general models. The cross-dataset search is also considered to develop more general architectures. Eventually, we design a cell cascaded pyramid network (C2PN) for final detection, which aggregates multiscale features for performance improvements. Compared with state-of-the-art networks artificially designed, our method achieves competitive performance in both in-dataset and cross-dataset scenarios.
4346.
Sarker, Prodip Kumar
Transformer-based neural architecture search for effective visible-infrared person re-identification Journal Article
In: Neurocomputing, vol. 620, pp. 129257, 2025, ISSN: 0925-2312.
@article{SARKER2025129257,
title = {Transformer-based neural architecture search for effective visible-infrared person re-identification},
author = {Prodip Kumar Sarker},
url = {https://www.sciencedirect.com/science/article/pii/S0925231224020289},
doi = {https://doi.org/10.1016/j.neucom.2024.129257},
issn = {0925-2312},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
journal = {Neurocomputing},
volume = {620},
pages = {129257},
abstract = {Visible-infrared person re-identification (VI-reID) is a complex task in security and video surveillance that aims to identify and match a person captured by various non-overlapping cameras. In recent years, there has been a notable advancement in reID owing to the development of transformer-based architectures. Although many existing methods emphasize on learning both modality-specific and shared features, challenges remain in fully exploiting the complementary information between infrared and visible modalities. Consequently, there is still opportunity to increase retrieval performance by effectively comprehending and integrating cross-modality semantic information. These designs often have problems with model complexity and time-consuming processes. To tackle these issues, we employ a novel transformer-based neural architecture search (TNAS) deep learning approach for effective VI-reID. To alleviate modality gaps, we first introduce a global–local transformer (GLT) module that captures features at both global and local levels across different modalities, contributing to better feature representation and matching. Then, an efficient neural architecture search (NAS) module is developed to search for the optimal transformer-based architecture, which further enhances the performance of VI-reID. Additionally, we introduce distillation loss and modality discriminative (MD) loss to examine the potential consistency between different modalities to promote intermodality separation between classes and intramodality compactness within classes. Experimental results on two challenging benchmark datasets illustrate that our developed model achieves state-of-the-art results, outperforming existing VI-reID methods.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Visible-infrared person re-identification (VI-reID) is a complex task in security and video surveillance that aims to identify and match a person captured by various non-overlapping cameras. In recent years, there has been a notable advancement in reID owing to the development of transformer-based architectures. Although many existing methods emphasize on learning both modality-specific and shared features, challenges remain in fully exploiting the complementary information between infrared and visible modalities. Consequently, there is still opportunity to increase retrieval performance by effectively comprehending and integrating cross-modality semantic information. These designs often have problems with model complexity and time-consuming processes. To tackle these issues, we employ a novel transformer-based neural architecture search (TNAS) deep learning approach for effective VI-reID. To alleviate modality gaps, we first introduce a global–local transformer (GLT) module that captures features at both global and local levels across different modalities, contributing to better feature representation and matching. Then, an efficient neural architecture search (NAS) module is developed to search for the optimal transformer-based architecture, which further enhances the performance of VI-reID. Additionally, we introduce distillation loss and modality discriminative (MD) loss to examine the potential consistency between different modalities to promote intermodality separation between classes and intramodality compactness within classes. Experimental results on two challenging benchmark datasets illustrate that our developed model achieves state-of-the-art results, outperforming existing VI-reID methods.
4345.
Zhang, Zhengzhuo; Zhuang, Liansheng
Progressive Neural Architecture Generation with Weaker Predictors Proceedings Article
In: Ide, Ichiro; Kompatsiaris, Ioannis; Xu, Changsheng; Yanai, Keiji; Chu, Wei-Ta; Nitta, Naoko; Riegler, Michael; Yamasaki, Toshihiko (Ed.): MultiMedia Modeling, pp. 229–242, Springer Nature Singapore, Singapore, 2025, ISBN: 978-981-96-2064-7.
@inproceedings{10.1007/978-981-96-2064-7_17,
title = {Progressive Neural Architecture Generation with Weaker Predictors},
author = {Zhengzhuo Zhang and Liansheng Zhuang},
editor = {Ichiro Ide and Ioannis Kompatsiaris and Changsheng Xu and Keiji Yanai and Wei-Ta Chu and Naoko Nitta and Michael Riegler and Toshihiko Yamasaki},
isbn = {978-981-96-2064-7},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
booktitle = {MultiMedia Modeling},
pages = {229–242},
publisher = {Springer Nature Singapore},
address = {Singapore},
abstract = {Neural Architecture Search (NAS) methods often suffer from low search efficiency since they have to explore a large and complex architecture search space. To accelerate the architecture search, generative methods learn a search space capturing intricate architecture distributions and generate promising architectures guided by a strong predictor within the latent space. However, since the architecture space is often exponentially large and highly non-convex, even a very strong predictor model is difficult in fitting the whole space, which may degrade their performance. To address this problem, this paper proposes a novel framework named Progressive Neural Architecture Generation with Weaker Predictors (WeakPNAG), which uses conditional diffusion models to generate promising architectures guided by weak predictors. Different from existing generative NAS methods which use a single strong predictor, our WeakPNAG progressively shrinks the sample space based on predictions from previous weak predictors, and updates new weak predictors towards the subspace of better architectures. In this way, our WeakPNAG iteratively learns to generate samples from increasingly promising latent subspaces. Extensive experiments on standard benchmarks demonstrate that our WeakPNAG achieves superior performance with reduced evaluation time compared with SOTA NAS methods.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Neural Architecture Search (NAS) methods often suffer from low search efficiency since they have to explore a large and complex architecture search space. To accelerate the architecture search, generative methods learn a search space capturing intricate architecture distributions and generate promising architectures guided by a strong predictor within the latent space. However, since the architecture space is often exponentially large and highly non-convex, even a very strong predictor model is difficult in fitting the whole space, which may degrade their performance. To address this problem, this paper proposes a novel framework named Progressive Neural Architecture Generation with Weaker Predictors (WeakPNAG), which uses conditional diffusion models to generate promising architectures guided by weak predictors. Different from existing generative NAS methods which use a single strong predictor, our WeakPNAG progressively shrinks the sample space based on predictions from previous weak predictors, and updates new weak predictors towards the subspace of better architectures. In this way, our WeakPNAG iteratively learns to generate samples from increasingly promising latent subspaces. Extensive experiments on standard benchmarks demonstrate that our WeakPNAG achieves superior performance with reduced evaluation time compared with SOTA NAS methods.
4344.
Su, Yulan; Zhang, Sisi; Lin, Zechao; Wang, Xingbin; Zhao, Lutan; Meng, Dan; Hou, Rui
Poseidon: A NAS-Based Ensemble Defense Method Against Multiple Perturbations Proceedings Article
In: Ide, Ichiro; Kompatsiaris, Ioannis; Xu, Changsheng; Yanai, Keiji; Chu, Wei-Ta; Nitta, Naoko; Riegler, Michael; Yamasaki, Toshihiko (Ed.): MultiMedia Modeling, pp. 215–228, Springer Nature Singapore, Singapore, 2025, ISBN: 978-981-96-2064-7.
@inproceedings{10.1007/978-981-96-2064-7_16,
title = {Poseidon: A NAS-Based Ensemble Defense Method Against Multiple Perturbations},
author = {Yulan Su and Sisi Zhang and Zechao Lin and Xingbin Wang and Lutan Zhao and Dan Meng and Rui Hou},
editor = {Ichiro Ide and Ioannis Kompatsiaris and Changsheng Xu and Keiji Yanai and Wei-Ta Chu and Naoko Nitta and Michael Riegler and Toshihiko Yamasaki},
isbn = {978-981-96-2064-7},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
booktitle = {MultiMedia Modeling},
pages = {215–228},
publisher = {Springer Nature Singapore},
address = {Singapore},
abstract = {Deep learning models have been proven to be severely affected by adversarial examples, which limit the widespread deployment of deep learning models. Prior research largely focused on defending against single types of perturbations using a single network. However, these methods are susceptible to sacrificing defense range because defense models that are trained to be robust against one perturbation type may not be robust against other types. Moreover, it is unrealistic to assume that neural networks would only be affected by a single type of perturbation. To defend against multiple perturbations, recent works have attempted to improve the overall robustness against multiple perturbations. Nonetheless, when evaluating the model's robustness against each type of perturbation, multi-perturbation defenses are still significantly less effective than models that are robust against a single perturbation type. To address these issues, we propose Poseidon, an ensemble defense method based on neural architecture search to defend against multiple perturbations. We first highlight the importance of architecture evolution in enhancing model robustness. And a novel robust architecture search method is proposed to identify perturbation-tailored architectures for sub-models. Furthermore, we explore a dedicated ensemble method that can combine these diverse sub-model architectures to be robust against multiple types of perturbations. The experimental results demonstrate that Poseidon outperforms the state-of-the-art multiple perturbation defense methods by 10.9% and 7.4% in robustness on the CIFAR-10 and CIFAR-100 datasets, respectively.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Deep learning models have been proven to be severely affected by adversarial examples, which limit the widespread deployment of deep learning models. Prior research largely focused on defending against single types of perturbations using a single network. However, these methods are susceptible to sacrificing defense range because defense models that are trained to be robust against one perturbation type may not be robust against other types. Moreover, it is unrealistic to assume that neural networks would only be affected by a single type of perturbation. To defend against multiple perturbations, recent works have attempted to improve the overall robustness against multiple perturbations. Nonetheless, when evaluating the model's robustness against each type of perturbation, multi-perturbation defenses are still significantly less effective than models that are robust against a single perturbation type. To address these issues, we propose Poseidon, an ensemble defense method based on neural architecture search to defend against multiple perturbations. We first highlight the importance of architecture evolution in enhancing model robustness. And a novel robust architecture search method is proposed to identify perturbation-tailored architectures for sub-models. Furthermore, we explore a dedicated ensemble method that can combine these diverse sub-model architectures to be robust against multiple types of perturbations. The experimental results demonstrate that Poseidon outperforms the state-of-the-art multiple perturbation defense methods by 10.9% and 7.4% in robustness on the CIFAR-10 and CIFAR-100 datasets, respectively.
4343.
Amoury, Sofia El; Smili, Youssef; Fakhri, Youssef
Design of an Optimal Convolutional Neural Network Architecture for MRI Brain Tumor Classification by Exploiting Particle Swarm Optimization Journal Article
In: Preprints, 2025.
@article{202501.0040,
title = {Design of an Optimal Convolutional Neural Network Architecture for MRI Brain Tumor Classification by Exploiting Particle Swarm Optimization},
author = {Sofia El Amoury and Youssef Smili and Youssef Fakhri},
url = {https://doi.org/10.20944/preprints202501.0040.v1},
doi = {10.20944/preprints202501.0040.v1},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
journal = {Preprints},
publisher = {Preprints},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
4342.
Capello, Alessio; Berta, Riccardo; Ballout, Hadi; Fresta, Matteo; Soltanmuradov, Vafali; Bellotti, Francesco
Enhancing μNAS for 1D CNNs on Microcontrollers Proceedings Article
In: Valle, Maurizio; Gastaldo, Paolo; Limiti, Ernesto (Ed.): Proceedings of SIE 2024, pp. 481–486, Springer Nature Switzerland, Cham, 2025, ISBN: 978-3-031-71518-1.
@inproceedings{10.1007/978-3-031-71518-1_59,
title = {Enhancing μNAS for 1D CNNs on Microcontrollers},
author = {Alessio Capello and Riccardo Berta and Hadi Ballout and Matteo Fresta and Vafali Soltanmuradov and Francesco Bellotti},
editor = {Maurizio Valle and Paolo Gastaldo and Ernesto Limiti},
url = {https://link.springer.com/chapter/10.1007/978-3-031-71518-1_59},
isbn = {978-3-031-71518-1},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
booktitle = {Proceedings of SIE 2024},
pages = {481–486},
publisher = {Springer Nature Switzerland},
address = {Cham},
abstract = {Deep Learning (DL) has proved effective in a variety of application domains. However, the computational and memory demand posed by deep neural models limits the spread of DL on resource-constrained devices such as microcontrollers. An opportunity to tailor DL models to specific hardware constraints is given by Neural Architecture Search (NAS), which automatically traverses a large search space, seeking for optimal architectures both in terms of hardware and performance, based on user specifications. State of the art open-source NAS tools for microcontrollers only support 2D Convolutional Neural Network (CNN) and Multi Layer Perceptron (MLP), but do not consider 1D convolution, which is key for time series analysis and signal processing. This study focuses on enhancing the state-of-the-art μNAS framework, by adding support for 1D CNN. Preliminary tests on a dummy dataset consisting of simple gaussian-distributed waveforms, demonstrate the system ability to find appropriate architectures to satisfy the specified constraints.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Deep Learning (DL) has proved effective in a variety of application domains. However, the computational and memory demand posed by deep neural models limits the spread of DL on resource-constrained devices such as microcontrollers. An opportunity to tailor DL models to specific hardware constraints is given by Neural Architecture Search (NAS), which automatically traverses a large search space, seeking for optimal architectures both in terms of hardware and performance, based on user specifications. State of the art open-source NAS tools for microcontrollers only support 2D Convolutional Neural Network (CNN) and Multi Layer Perceptron (MLP), but do not consider 1D convolution, which is key for time series analysis and signal processing. This study focuses on enhancing the state-of-the-art μNAS framework, by adding support for 1D CNN. Preliminary tests on a dummy dataset consisting of simple gaussian-distributed waveforms, demonstrate the system ability to find appropriate architectures to satisfy the specified constraints.
4341.
Xu, Jingjing; Wu, Caesar; Li, Yuan-Fang; Danoy, Grégoire; Bouvry, Pascal
A Unified Hyperparameter Optimization Pipeline for Transformer-Based Time Series Forecasting Models Miscellaneous
2025.
@misc{xu2025unifiedhyperparameteroptimizationpipeline,
title = {A Unified Hyperparameter Optimization Pipeline for Transformer-Based Time Series Forecasting Models},
author = {Jingjing Xu and Caesar Wu and Yuan-Fang Li and Grégoire Danoy and Pascal Bouvry},
url = {https://arxiv.org/abs/2501.01394},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
keywords = {},
pubstate = {published},
tppubtype = {misc}
}
4340.
Ge, Chendi; Wang, Xin; Zhang, Ziwei; Qin, Yijian; Wu, Haiyang; Zhang, Yang; Yang, Yuekui; Zhu, Wenwu
Behavior Importance-Aware Graph Neural Architecture Search for Cross-Domain Recommendation Miscellaneous
2025.
@misc{nokey,
title = {Behavior Importance-Aware Graph Neural Architecture Search for Cross-Domain Recommendation},
author = {Chendi Ge and Xin Wang and Ziwei Zhang and Yijian Qin and Haiyang Wu and Yang Zhang and Yuekui Yang and Wenwu Zhu},
url = {https://mn.cs.tsinghua.edu.cn/xinwang/PDF/papers/2025_Behavior%20Importance-Aware%20Graph%20Neural%20Architecture%20Search%20for%20Cross-Domain%20Recommendation.pdf},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
keywords = {},
pubstate = {published},
tppubtype = {misc}
}
4339.
Wang, Mingzi; Meng, Yuan; Tang, Chen; Zhang, Weixiang; Qin, Yijian; Yao, Yang; Li, Yingxin; Feng, Tongtong; Wang, Xin; Guan, Xun; Wang, Zhi; Zhu, Wenwu
JAQ: Joint Efficient Architecture Design and Low-Bit Quantization with Hardware-Software Co-Exploration Miscellaneous
2025.
@misc{nokey,
title = {JAQ: Joint Efficient Architecture Design and Low-Bit Quantization with Hardware-Software Co-Exploration},
author = {Mingzi Wang and Yuan Meng and Chen Tang and Weixiang Zhang and Yijian Qin and Yang Yao and Yingxin Li and Tongtong Feng and Xin Wang and Xun Guan and Zhi Wang and Wenwu Zhu},
url = {https://mn.cs.tsinghua.edu.cn/xinwang/PDF/papers/2025_JAQ%20Joint%20Efficient%20Architecture%20Design%20and%20Low-Bit%20Quantization%20with%20Hardware-Software%20Co-Exploration.pdf},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
keywords = {},
pubstate = {published},
tppubtype = {misc}
}
4338.
Avval, Sasan Salmani Pour; Eskue, Nathan D.; Groves, Roger M.; Yaghoubi, Vahid
Systematic review on neural architecture search Journal Article
In: Artificial Intelligence Review , 2025.
@article{Avval-air25a,
title = {Systematic review on neural architecture search},
author = {Sasan Salmani Pour Avval and Nathan D. Eskue and Roger M. Groves and Vahid Yaghoubi
},
url = {https://link.springer.com/article/10.1007/s10462-024-11058-w},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
journal = {Artificial Intelligence Review },
keywords = {},
pubstate = {published},
tppubtype = {article}
}
4337.
He, Yannis Y.
TART: Token-based Architecture Transformer for Neural Network Performance Prediction Technical Report
2025.
@techreport{he2025tarttokenbasedarchitecturetransformer,
title = {TART: Token-based Architecture Transformer for Neural Network Performance Prediction},
author = {Yannis Y. He},
url = {https://arxiv.org/abs/2501.02007},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
keywords = {},
pubstate = {published},
tppubtype = {techreport}
}
4336.
Archet, Agathe; Ventroux, Nicolas; Gac, Nicolas; Orieux, François
A practical HW-aware NAS flow for AI vision applications on embedded heterogeneous SoCs Proceedings Article
In: International Workshop on Design and Architecture for Signal and Image Processing 2025, Barcelone, Spain, 2025.
@inproceedings{archet:hal-04869471,
title = {A practical HW-aware NAS flow for AI vision applications on embedded heterogeneous SoCs},
author = {Agathe Archet and Nicolas Ventroux and Nicolas Gac and François Orieux},
url = {https://hal.science/hal-04869471},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
booktitle = {International Workshop on Design and Architecture for Signal and Image Processing 2025},
address = {Barcelone, Spain},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
4335.
Luo, Zhirui; Li, Qingqing; Qi, Ruobin; Zheng, Jun
In: AI, vol. 6, no. 1, 2025, ISSN: 2673-2688.
@article{ai6010009,
title = {Designing Channel Attention Fully Convolutional Networks with Neural Architecture Search for Customer Socio-Demographic Information Identification Using Smart Meter Data},
author = {Zhirui Luo and Qingqing Li and Ruobin Qi and Jun Zheng},
url = {https://www.mdpi.com/2673-2688/6/1/9},
doi = {10.3390/ai6010009},
issn = {2673-2688},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
journal = {AI},
volume = {6},
number = {1},
abstract = {Background: Accurately identifying the socio-demographic information of customers is crucial for utilities. It enables them to efficiently deliver personalized energy services and manage distribution networks. In recent years, machine learning-based data-driven methods have gained popularity compared to traditional survey-based approaches, owing to their time and cost efficiency, as well as the availability of a large amount of high-frequency smart meter data. Methods: In this paper, we propose a new method that harnesses the power of neural architecture search to automatically design deep neural network architectures tailored for identifying various socio-demographic information of customers using smart meter data. We designed a search space based on a novel channel attention fully convolutional network architecture. Furthermore, we developed a search algorithm based on Bayesian optimization to effectively explore the space and identify high-performing architectures. Results: The performance of the proposed method was evaluated and compared with a set of machine learning and deep learning baseline methods using a smart meter dataset widely used in this research area. Our results show that the deep neural network architectures designed automatically by our proposed method significantly outperform all baseline methods in addressing the socio-demographic questions investigated in our study.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Background: Accurately identifying the socio-demographic information of customers is crucial for utilities. It enables them to efficiently deliver personalized energy services and manage distribution networks. In recent years, machine learning-based data-driven methods have gained popularity compared to traditional survey-based approaches, owing to their time and cost efficiency, as well as the availability of a large amount of high-frequency smart meter data. Methods: In this paper, we propose a new method that harnesses the power of neural architecture search to automatically design deep neural network architectures tailored for identifying various socio-demographic information of customers using smart meter data. We designed a search space based on a novel channel attention fully convolutional network architecture. Furthermore, we developed a search algorithm based on Bayesian optimization to effectively explore the space and identify high-performing architectures. Results: The performance of the proposed method was evaluated and compared with a set of machine learning and deep learning baseline methods using a smart meter dataset widely used in this research area. Our results show that the deep neural network architectures designed automatically by our proposed method significantly outperform all baseline methods in addressing the socio-demographic questions investigated in our study.
4334.
Yu, Caiyang; Wang, Jian; Wang, Yifan; Ju, Wei; Tang, Chenwei; Lv, Jiancheng
Rethinking neural architecture representation for predictors: Topological encoding in pixel space Journal Article
In: Information Fusion, vol. 118, pp. 102925, 2025, ISSN: 1566-2535.
@article{YU2025102925,
title = {Rethinking neural architecture representation for predictors: Topological encoding in pixel space},
author = {Caiyang Yu and Jian Wang and Yifan Wang and Wei Ju and Chenwei Tang and Jiancheng Lv},
url = {https://www.sciencedirect.com/science/article/pii/S1566253524007036},
doi = {https://doi.org/10.1016/j.inffus.2024.102925},
issn = {1566-2535},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
journal = {Information Fusion},
volume = {118},
pages = {102925},
abstract = {Neural predictors (NPs) aim to swiftly evaluate architectures during the neural architecture search (NAS) process. Precise evaluations with NPs heavily depend on the representation of training samples (i.e., the architectures), as the representation determines how well the NP captures the intrinsic properties and intricate dependencies of the architecture. Existing methods, which represent neural architectures as graph structures or sequences, are inherently limited in their expressive capabilities. In this study, we explore the image representation of neural architecture, describing the architecture in pixel space and using the long-range modeling capability of attention mechanisms to construct connections among pixels and extract tangible (tractable) architecture topology representation from them. Our attempt provides an efficient architecture representation for NPs, combined with today’s powerful pre-training models, showing promising prospects. Furthermore, recognizing that images alone may fall short in capturing configuration specifics, we design a corresponding text representation to provide a more accurate complement. Our experimental analysis reveals that the existing visual language model can efficiently identify the topological information in the pixel space. Additionally, we propose a Dual-Input Multichannel Neural Predictor (DIMNP) that simultaneously accepts multiple representations of architectures, facilitating information complementarity and accelerating convergence of the NP. Extensive experiments on NAS-Bench-101, NAS-Bench-201, and DARTS datasets demonstrate the superiority of DIMNP compared to the state-of-the-art NPs. In particular, on the NAS-Bench-101 and NAS-Bench-201 search spaces, DIMNP achieves performance improvements of 0.01 and 0.52, respectively, compared to the second-best algorithm on average.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Neural predictors (NPs) aim to swiftly evaluate architectures during the neural architecture search (NAS) process. Precise evaluations with NPs heavily depend on the representation of training samples (i.e., the architectures), as the representation determines how well the NP captures the intrinsic properties and intricate dependencies of the architecture. Existing methods, which represent neural architectures as graph structures or sequences, are inherently limited in their expressive capabilities. In this study, we explore the image representation of neural architecture, describing the architecture in pixel space and using the long-range modeling capability of attention mechanisms to construct connections among pixels and extract tangible (tractable) architecture topology representation from them. Our attempt provides an efficient architecture representation for NPs, combined with today’s powerful pre-training models, showing promising prospects. Furthermore, recognizing that images alone may fall short in capturing configuration specifics, we design a corresponding text representation to provide a more accurate complement. Our experimental analysis reveals that the existing visual language model can efficiently identify the topological information in the pixel space. Additionally, we propose a Dual-Input Multichannel Neural Predictor (DIMNP) that simultaneously accepts multiple representations of architectures, facilitating information complementarity and accelerating convergence of the NP. Extensive experiments on NAS-Bench-101, NAS-Bench-201, and DARTS datasets demonstrate the superiority of DIMNP compared to the state-of-the-art NPs. In particular, on the NAS-Bench-101 and NAS-Bench-201 search spaces, DIMNP achieves performance improvements of 0.01 and 0.52, respectively, compared to the second-best algorithm on average.
4333.
Yu, Jiandong; Li, Tongtong; Shi, Xuerong; Zhao, Ziyang; Chen, Miao; Zhang, Yu; Wang, Junyu; Yao, Zhijun; Fang, Lei; Hu, Bin
ETMO-NAS: An efficient two-step multimodal one-shot NAS for lung nodules classification Journal Article
In: Biomedical Signal Processing and Control, vol. 104, pp. 107479, 2025, ISSN: 1746-8094.
@article{YU2025107479,
title = {ETMO-NAS: An efficient two-step multimodal one-shot NAS for lung nodules classification},
author = {Jiandong Yu and Tongtong Li and Xuerong Shi and Ziyang Zhao and Miao Chen and Yu Zhang and Junyu Wang and Zhijun Yao and Lei Fang and Bin Hu},
url = {https://www.sciencedirect.com/science/article/pii/S1746809424015374},
doi = {https://doi.org/10.1016/j.bspc.2024.107479},
issn = {1746-8094},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
journal = {Biomedical Signal Processing and Control},
volume = {104},
pages = {107479},
abstract = {Malignant lung nodules are the initial diagnostic manifestation of lung cancer. Accurate predictive classification of malignant from benign lung nodules can improve treatment efficacy and survival rate of lung cancer patients. Since current deep learning-based PET/CT pulmonary nodule-assisted diagnosis models typically rely on network architectures carefully designed by researchers, which require professional expertise and extensive prior knowledge. To combat these challenges, in this paper, we propose an efficient two-step multimodal one-shot NAS (ETMO-NAS) for searching high-performance network architectures for reliable and accurate classification of lung nodules for multimodal PET/CT data. Specifically, the step I focuses on fully training the performance of all candidate architectures in the search space using the sandwich rule and in-place distillation strategy. The step II aims to split the search space into multiple non-overlapping subsupernets by parallel operation edge decomposition strategy and then fine-tune the subsupernets further improve performance. Finally, the performance of ETMO-NAS was validated on a set of real clinical data. The experimental results show that the classification architecture searched by ETMO-NAS achieves the best performance with accuracy, precision, sensitivity, specificity, and F-1 score of 94.23%, 92.10%, 95.83%, 92.86% and 0.9388, respectively. In addition, compared with the classical CNN model and NAS model, ETMO-NAS performs better with the same inputs, but with only 1/33–1/5 of the parameters. This provides substantial evidence for the competitiveness of the model in classification tasks and presents a new approach for automated diagnosis of PET/CT pulmonary nodules. Code and models will be available at: https://github.com/yujiandong0002/ETMO-NAS.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Malignant lung nodules are the initial diagnostic manifestation of lung cancer. Accurate predictive classification of malignant from benign lung nodules can improve treatment efficacy and survival rate of lung cancer patients. Since current deep learning-based PET/CT pulmonary nodule-assisted diagnosis models typically rely on network architectures carefully designed by researchers, which require professional expertise and extensive prior knowledge. To combat these challenges, in this paper, we propose an efficient two-step multimodal one-shot NAS (ETMO-NAS) for searching high-performance network architectures for reliable and accurate classification of lung nodules for multimodal PET/CT data. Specifically, the step I focuses on fully training the performance of all candidate architectures in the search space using the sandwich rule and in-place distillation strategy. The step II aims to split the search space into multiple non-overlapping subsupernets by parallel operation edge decomposition strategy and then fine-tune the subsupernets further improve performance. Finally, the performance of ETMO-NAS was validated on a set of real clinical data. The experimental results show that the classification architecture searched by ETMO-NAS achieves the best performance with accuracy, precision, sensitivity, specificity, and F-1 score of 94.23%, 92.10%, 95.83%, 92.86% and 0.9388, respectively. In addition, compared with the classical CNN model and NAS model, ETMO-NAS performs better with the same inputs, but with only 1/33–1/5 of the parameters. This provides substantial evidence for the competitiveness of the model in classification tasks and presents a new approach for automated diagnosis of PET/CT pulmonary nodules. Code and models will be available at: https://github.com/yujiandong0002/ETMO-NAS.
4332.
Saeed, Farah; Tan, Chenjiao; Liu, Tianming; Li, Changying
3D neural architecture search to optimize segmentation of plant parts Journal Article
In: Smart Agricultural Technology, vol. 10, pp. 100776, 2025, ISSN: 2772-3755.
@article{SAEED2025100776,
title = {3D neural architecture search to optimize segmentation of plant parts},
author = {Farah Saeed and Chenjiao Tan and Tianming Liu and Changying Li},
url = {https://www.sciencedirect.com/science/article/pii/S2772375525000103},
doi = {https://doi.org/10.1016/j.atech.2025.100776},
issn = {2772-3755},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
journal = {Smart Agricultural Technology},
volume = {10},
pages = {100776},
abstract = {Accurately segmenting plant parts from imagery is vital for improving crop phenotypic traits. However, current 3D deep learning models for segmentation in point cloud data require specific network architectures that are usually manually designed, which is both tedious and suboptimal. To overcome this issue, a 3D neural architecture search (NAS) was performed in this study to optimize cotton plant part segmentation. The search space was designed using Point Voxel Convolution (PVConv) as the basic building block of the network. The NAS framework included a supernetwork with weight sharing and an evolutionary search to find optimal candidates, with three surrogate learners to predict mean IoU, latency, and memory footprint. The optimal candidate searched from the proposed method consisted of five PVConv layers with either 32 or 512 output channels, achieving mean IoU and accuracy of over 90 % and 96 %, respectively, and outperforming manually designed architectures. Additionally, the evolutionary search was updated to search for architectures satisfying memory and time constraints, with searched architectures achieving mean IoU and accuracy of >84 % and 94 %, respectively. Furthermore, a differentiable architecture search (DARTS) utilizing PVConv operation was implemented for comparison, and our method demonstrated better segmentation performance with a margin of >2 % and 1 % in mean IoU and accuracy, respectively. Overall, the proposed method can be applied to segment cotton plants with an accuracy over 94 %, while adjusting to available resource constraints.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Accurately segmenting plant parts from imagery is vital for improving crop phenotypic traits. However, current 3D deep learning models for segmentation in point cloud data require specific network architectures that are usually manually designed, which is both tedious and suboptimal. To overcome this issue, a 3D neural architecture search (NAS) was performed in this study to optimize cotton plant part segmentation. The search space was designed using Point Voxel Convolution (PVConv) as the basic building block of the network. The NAS framework included a supernetwork with weight sharing and an evolutionary search to find optimal candidates, with three surrogate learners to predict mean IoU, latency, and memory footprint. The optimal candidate searched from the proposed method consisted of five PVConv layers with either 32 or 512 output channels, achieving mean IoU and accuracy of over 90 % and 96 %, respectively, and outperforming manually designed architectures. Additionally, the evolutionary search was updated to search for architectures satisfying memory and time constraints, with searched architectures achieving mean IoU and accuracy of >84 % and 94 %, respectively. Furthermore, a differentiable architecture search (DARTS) utilizing PVConv operation was implemented for comparison, and our method demonstrated better segmentation performance with a margin of >2 % and 1 % in mean IoU and accuracy, respectively. Overall, the proposed method can be applied to segment cotton plants with an accuracy over 94 %, while adjusting to available resource constraints.
4331.
Weitz, Jason; Demler, Dmitri; McDermott, Luke; Tran, Nhan; Duarte, Javier
Neural Architecture Codesign for Fast Physics Applications Technical Report
2025.
@techreport{weitz2025neuralarchitecturecodesignfast,
title = {Neural Architecture Codesign for Fast Physics Applications},
author = {Jason Weitz and Dmitri Demler and Luke McDermott and Nhan Tran and Javier Duarte},
url = {https://arxiv.org/abs/2501.05515},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
keywords = {},
pubstate = {published},
tppubtype = {techreport}
}
4330.
Song, Xiaotian; Lv, Zeqiong; Fan, Jiaohao; Deng, Xiong; Lv, Jiancheng; Liu, Jiyuan; Sun, Yanan
Evolutionary Multi-Objective Spiking Neural Architecture Search for Image Classification Journal Article
In: IEEE Transactions on Evolutionary Computation, pp. 1-1, 2025.
@article{10838601,
title = {Evolutionary Multi-Objective Spiking Neural Architecture Search for Image Classification},
author = {Xiaotian Song and Zeqiong Lv and Jiaohao Fan and Xiong Deng and Jiancheng Lv and Jiyuan Liu and Yanan Sun},
doi = {10.1109/TEVC.2025.3528471},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
journal = {IEEE Transactions on Evolutionary Computation},
pages = {1-1},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
4329.
Louati, Hassen; Louati, Ali; Mansour, Khalid; Kariri, Elham
Achieving Faster and Smarter Chest X-Ray Classification With Optimized CNNs Journal Article
In: IEEE Access, vol. 13, pp. 10070–10082, 2025.
@article{DBLP:journals/access/LouatiLMK25,
title = {Achieving Faster and Smarter Chest X-Ray Classification With Optimized CNNs},
author = {Hassen Louati and Ali Louati and Khalid Mansour and Elham Kariri},
url = {https://doi.org/10.1109/ACCESS.2025.3529206},
doi = {10.1109/ACCESS.2025.3529206},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
journal = {IEEE Access},
volume = {13},
pages = {10070–10082},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
4328.
Nie, Mingshuo; Chen, Dongming; Chen, Huilin; Wang, Dongqi
AutoMTNAS: Automated meta-reinforcement learning on graph tokenization for graph neural architecture search Journal Article
In: Knowledge-Based Systems, vol. 310, pp. 113023, 2025, ISSN: 0950-7051.
@article{NIE2025113023,
title = {AutoMTNAS: Automated meta-reinforcement learning on graph tokenization for graph neural architecture search},
author = {Mingshuo Nie and Dongming Chen and Huilin Chen and Dongqi Wang},
url = {https://www.sciencedirect.com/science/article/pii/S0950705125000711},
doi = {https://doi.org/10.1016/j.knosys.2025.113023},
issn = {0950-7051},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
journal = {Knowledge-Based Systems},
volume = {310},
pages = {113023},
abstract = {Graph neural networks have achieved breakthroughs in various fields due to their powerful automated representation capabilities for graph. Designing effective graph neural architectures is critical for feature representation and property prediction in non-Euclidean graph-structured data. However, this design process heavily relies on the strong prior knowledge and experience of researchers. The inherent complexity and irregularity in graph-structured data make it challenging for existing methods to develop strategies for capturing expressive representations beyond traditional paradigms, resulting in unaffordable computational cost and precision loss across diverse graphs. To this end, we propose a novel automated meta-reinforcement learning on graph tokenization for graph neural architecture search, named AutoMTNAS, to learn a more general and reliable architecture search policy. In particular, our graph tokenization method identifies critical nodes and structural patterns within the graph and captures label-aware global information to summarize potential valuable insights. We define a meta-reinforcement learning searcher that utilizes parameter sharing and policy gradients to discover optimal architectures for new tasks, even with limited available observations. Extensive experiments on benchmark datasets, ranging from small to large, demonstrate that AutoMTNAS outperforms human-invented architectures and existing graph neural architecture search methods.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Graph neural networks have achieved breakthroughs in various fields due to their powerful automated representation capabilities for graph. Designing effective graph neural architectures is critical for feature representation and property prediction in non-Euclidean graph-structured data. However, this design process heavily relies on the strong prior knowledge and experience of researchers. The inherent complexity and irregularity in graph-structured data make it challenging for existing methods to develop strategies for capturing expressive representations beyond traditional paradigms, resulting in unaffordable computational cost and precision loss across diverse graphs. To this end, we propose a novel automated meta-reinforcement learning on graph tokenization for graph neural architecture search, named AutoMTNAS, to learn a more general and reliable architecture search policy. In particular, our graph tokenization method identifies critical nodes and structural patterns within the graph and captures label-aware global information to summarize potential valuable insights. We define a meta-reinforcement learning searcher that utilizes parameter sharing and policy gradients to discover optimal architectures for new tasks, even with limited available observations. Extensive experiments on benchmark datasets, ranging from small to large, demonstrate that AutoMTNAS outperforms human-invented architectures and existing graph neural architecture search methods.
4327.
Jiang, Pengcheng; Xue, Yu; Neri, Ferrante
Score Predictor-Assisted Evolutionary Neural Architecture Search Journal Article
In: IEEE Transactions on Emerging Topics in Computational Intelligence, pp. 1-15, 2025.
@article{10841460,
title = {Score Predictor-Assisted Evolutionary Neural Architecture Search},
author = {Pengcheng Jiang and Yu Xue and Ferrante Neri},
url = {https://ieeexplore.ieee.org/abstract/document/10841460},
doi = {10.1109/TETCI.2025.3526179},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
journal = {IEEE Transactions on Emerging Topics in Computational Intelligence},
pages = {1-15},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
4326.
Slimani, Hicham; Mhamdi, Jamal El; Jilbab, Abdelilah
Advanced Algorithmic Model for Real-Time Multi-Level Crop Disease Detection Using Neural Architecture Search Journal Article
In: E3S Web Conf., vol. 601, pp. 00032, 2025.
@article{refId0c,
title = {Advanced Algorithmic Model for Real-Time Multi-Level Crop Disease Detection Using Neural Architecture Search},
author = {Hicham Slimani and Jamal El Mhamdi and Abdelilah Jilbab},
url = {https://doi.org/10.1051/e3sconf/202560100032},
doi = {10.1051/e3sconf/202560100032},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
journal = {E3S Web Conf.},
volume = {601},
pages = {00032},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
4325.
Abebe, Waqwoya; Jafari, Sadegh; Yu, Sixing; Dutta, Akash; Strube, Jan; Tallent, Nathan R.; Guo, Luanzheng; Munoz, Pablo; Jannesari, Ali
SuperSAM: Crafting a SAM Supernetwork via Structured Pruning and Unstructured Parameter Prioritization Technical Report
2025.
@techreport{abebe2025supersamcraftingsamsupernetwork,
title = {SuperSAM: Crafting a SAM Supernetwork via Structured Pruning and Unstructured Parameter Prioritization},
author = {Waqwoya Abebe and Sadegh Jafari and Sixing Yu and Akash Dutta and Jan Strube and Nathan R. Tallent and Luanzheng Guo and Pablo Munoz and Ali Jannesari},
url = {https://arxiv.org/abs/2501.08504},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
keywords = {},
pubstate = {published},
tppubtype = {techreport}
}
4324.
Youm, Sungkwan; Go, Sunghyun
Lightweight and Efficient CSI-Based Human Activity Recognition via Bayesian Optimization-Guided Architecture Search and Structured Pruning Journal Article
In: Applied Sciences, vol. 15, no. 2, 2025, ISSN: 2076-3417.
@article{app15020890,
title = {Lightweight and Efficient CSI-Based Human Activity Recognition via Bayesian Optimization-Guided Architecture Search and Structured Pruning},
author = {Sungkwan Youm and Sunghyun Go},
url = {https://www.mdpi.com/2076-3417/15/2/890},
doi = {10.3390/app15020890},
issn = {2076-3417},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
journal = {Applied Sciences},
volume = {15},
number = {2},
abstract = {This paper presents an integrated approach to developing lightweight, high-performance deep learning models for human activity recognition (HAR) using WiFi Channel State Information (CSI). Motivated by the need for accuracy and efficiency in resource-constrained environments, we combine Bayesian Optimization-based Neural Architecture Search (NAS) with a structured pruning algorithm. NAS identifies optimal network configurations, while pruning systematically removes redundant parameters, preserving accuracy. This approach allows for robust activity recognition from diverse WiFi datasets under varying conditions. Experimental results across multiple benchmark datasets demonstrate that our method not only maintains but often improves accuracy after pruning, resulting in models that are both smaller and more accurate. This offers a scalable and adaptable solution for real-world deployments in IoT and mobile platforms, achieving an optimal balance of efficiency and accuracy in HAR using WiFi CSI.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
This paper presents an integrated approach to developing lightweight, high-performance deep learning models for human activity recognition (HAR) using WiFi Channel State Information (CSI). Motivated by the need for accuracy and efficiency in resource-constrained environments, we combine Bayesian Optimization-based Neural Architecture Search (NAS) with a structured pruning algorithm. NAS identifies optimal network configurations, while pruning systematically removes redundant parameters, preserving accuracy. This approach allows for robust activity recognition from diverse WiFi datasets under varying conditions. Experimental results across multiple benchmark datasets demonstrate that our method not only maintains but often improves accuracy after pruning, resulting in models that are both smaller and more accurate. This offers a scalable and adaptable solution for real-world deployments in IoT and mobile platforms, achieving an optimal balance of efficiency and accuracy in HAR using WiFi CSI.
4323.
Wu, Yue; Gong, Peiran; Yuan, Yongzhe; Gong, Maoguo; Ma, Wenping; Miao, Qiguang
Evolutionary Neural Architecture Search Framework With Masked Encoding Mechanism for Point Cloud Registration Journal Article
In: IEEE Transactions on Emerging Topics in Computational Intelligence, pp. 1-13, 2025.
@article{10847916,
title = {Evolutionary Neural Architecture Search Framework With Masked Encoding Mechanism for Point Cloud Registration},
author = {Yue Wu and Peiran Gong and Yongzhe Yuan and Maoguo Gong and Wenping Ma and Qiguang Miao},
url = {https://ieeexplore.ieee.org/abstract/document/10847916},
doi = {10.1109/TETCI.2025.3526279},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
journal = {IEEE Transactions on Emerging Topics in Computational Intelligence},
pages = {1-13},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
4322.
Zada, Moustafa
Hybrid-Quantum Neural Architecture Search for The Proximal Policy Optimization Algorithm Technical Report
2025.
@techreport{https://doi.org/10.5281/zenodo.14625856,
title = {Hybrid-Quantum Neural Architecture Search for The Proximal Policy Optimization Algorithm},
author = {Moustafa Zada},
url = {https://zenodo.org/doi/10.5281/zenodo.14625856},
doi = {10.5281/ZENODO.14625856},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
publisher = {Zenodo},
keywords = {},
pubstate = {published},
tppubtype = {techreport}
}
4321.
Douka, Stella; Verbockhaven, Manon; Rudkiewicz, Théo; Rivaud, Stéphane; Landes, François P; Chevallier, Sylvain; Charpiat, Guillaume
Growth strategies for arbitrary DAG neural architectures Technical Report
2025.
@techreport{douka2025growthstrategiesarbitrarydag,
title = {Growth strategies for arbitrary DAG neural architectures},
author = {Stella Douka and Manon Verbockhaven and Théo Rudkiewicz and Stéphane Rivaud and François P Landes and Sylvain Chevallier and Guillaume Charpiat},
url = {https://arxiv.org/abs/2501.12690},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
keywords = {},
pubstate = {published},
tppubtype = {techreport}
}
4320.
Deutel, Mark; Kontes, Georgios; Mutschler, Christopher; Teich, Jürgen
Combining Multi-Objective Bayesian Optimization with Reinforcement Learning for TinyML Journal Article
In: ACM Trans. Evol. Learn. Optim., 2025, (Just Accepted).
@article{10.1145/3715012,
title = {Combining Multi-Objective Bayesian Optimization with Reinforcement Learning for TinyML},
author = {Mark Deutel and Georgios Kontes and Christopher Mutschler and Jürgen Teich},
url = {https://doi.org/10.1145/3715012},
doi = {10.1145/3715012},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
journal = {ACM Trans. Evol. Learn. Optim.},
publisher = {Association for Computing Machinery},
address = {New York, NY, USA},
abstract = {Deploying deep neural networks (DNNs) on microcontrollers (TinyML) is a common trend to process the increasing amount of sensor data generated at the edge, but in practice, resource and latency constraints make it difficult to find optimal DNN candidates. Neural architecture search (NAS) is an excellent approach to automate this search and can easily be combined with DNN compression techniques commonly used in TinyML. However, many NAS techniques are not only computationally expensive, especially hyperparameter optimization (HPO), but also often focus on optimizing only a single objective, e.g., maximizing accuracy, without considering additional objectives such as memory requirements or computational complexity of a DNN, which are key to making deployment at the edge feasible. In this paper, we propose a novel NAS strategy for TinyML based on multi-objective Bayesian optimization (MOBOpt) and an ensemble of competing parametric policies trained using Augmented Random Search (ARS) reinforcement learning (RL) agents. Our methodology aims at efficiently finding tradeoffs between a DNN’s predictive accuracy, memory requirements on a given target system, and computational complexity. Our experiments show that we consistently outperform existing MOBOpt approaches on different datasets and architectures such as ResNet-18 and MobileNetV3.},
note = {Just Accepted},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Deploying deep neural networks (DNNs) on microcontrollers (TinyML) is a common trend to process the increasing amount of sensor data generated at the edge, but in practice, resource and latency constraints make it difficult to find optimal DNN candidates. Neural architecture search (NAS) is an excellent approach to automate this search and can easily be combined with DNN compression techniques commonly used in TinyML. However, many NAS techniques are not only computationally expensive, especially hyperparameter optimization (HPO), but also often focus on optimizing only a single objective, e.g., maximizing accuracy, without considering additional objectives such as memory requirements or computational complexity of a DNN, which are key to making deployment at the edge feasible. In this paper, we propose a novel NAS strategy for TinyML based on multi-objective Bayesian optimization (MOBOpt) and an ensemble of competing parametric policies trained using Augmented Random Search (ARS) reinforcement learning (RL) agents. Our methodology aims at efficiently finding tradeoffs between a DNN’s predictive accuracy, memory requirements on a given target system, and computational complexity. Our experiments show that we consistently outperform existing MOBOpt approaches on different datasets and architectures such as ResNet-18 and MobileNetV3.
4319.
Bastami, Sajad; Dolatshahi, Mohammad Bagher
Compact Neural Architecture Search for Image Classification Using Gravitational Search Algorithm Journal Article
In: Applied and basic Machine intelligence research, pp. 77-91, 2025, ISSN: 2821-2029.
@article{nokey,
title = {Compact Neural Architecture Search for Image Classification Using Gravitational Search Algorithm},
author = {Sajad Bastami and Mohammad Bagher Dolatshahi},
url = {https://abmir.yazd.ac.ir/article_3644.html},
doi = {10.22034/abmir.2024.22228.1066},
issn = {2821-2029},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
journal = {Applied and basic Machine intelligence research},
pages = {77-91},
publisher = {Yazd University},
abstract = {This paper presents a compact neural architecture search method for image classification using the Gravitational Search Algorithm (GSA). Deep learning, through multi-layer computational models, enables automatic feature extraction from raw data at various levels of abstraction, playing a key role in complex tasks such as image classification. Neural Architecture Search (NAS), which automatically discovers new architectures for Convolutional Neural Networks (CNNs), faces challenges such as high computational complexity and costs. To address these issues, a GSA-based approach has been developed, employing a bi-level variable-length optimization technique to design both micro and macro architectures of CNNs. This approach, leveraging a compact search space and modified convolutional bottlenecks, demonstrates superior performance compared to state-of-the-art methods. Experimental results on CIFAR-10, CIFAR-100, and ImageNet datasets reveal that the proposed method achieves a classification accuracy of 98.48% with a search cost of 1.05 GPU days, outperforming existing algorithms in terms of accuracy, search efficiency, and architectural complexity.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
This paper presents a compact neural architecture search method for image classification using the Gravitational Search Algorithm (GSA). Deep learning, through multi-layer computational models, enables automatic feature extraction from raw data at various levels of abstraction, playing a key role in complex tasks such as image classification. Neural Architecture Search (NAS), which automatically discovers new architectures for Convolutional Neural Networks (CNNs), faces challenges such as high computational complexity and costs. To address these issues, a GSA-based approach has been developed, employing a bi-level variable-length optimization technique to design both micro and macro architectures of CNNs. This approach, leveraging a compact search space and modified convolutional bottlenecks, demonstrates superior performance compared to state-of-the-art methods. Experimental results on CIFAR-10, CIFAR-100, and ImageNet datasets reveal that the proposed method achieves a classification accuracy of 98.48% with a search cost of 1.05 GPU days, outperforming existing algorithms in terms of accuracy, search efficiency, and architectural complexity.
4318.
Frachon, Luc J.
Novel approaches in macro-level neural ensemble architecture search PhD Thesis
2025.
@phdthesis{nokey,
title = {Novel approaches in macro-level neural ensemble architecture search},
author = {Frachon, Luc J.},
url = {https://www.ros.hw.ac.uk/handle/10399/5040},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
keywords = {},
pubstate = {published},
tppubtype = {phdthesis}
}
4317.
Ouertatani, H.; Maxim, C.; Niar, S.; Talbi, E-G.
Neural Architecture Tuning: A BO-Powered NAS Tool Proceedings Article
In: Dorronsoro, Bernabé; Zagar, Martin; Talbi, El-Ghazali (Ed.): Optimization and Learning, pp. 82–93, Springer Nature Switzerland, Cham, 2025, ISBN: 978-3-031-77941-1.
@inproceedings{10.1007/978-3-031-77941-1_7,
title = {Neural Architecture Tuning: A BO-Powered NAS Tool},
author = {H. Ouertatani and C. Maxim and S. Niar and E-G. Talbi},
editor = {Bernabé Dorronsoro and Martin Zagar and El-Ghazali Talbi},
url = {https://link.springer.com/chapter/10.1007/978-3-031-77941-1_7},
isbn = {978-3-031-77941-1},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
booktitle = {Optimization and Learning},
pages = {82–93},
publisher = {Springer Nature Switzerland},
address = {Cham},
abstract = {Neural Architecture Search (NAS) consists of applying an optimization technique to find the best performing architecture(s) in a defined search space, with regard to an objective function. The practical implementation of NAS currently carries certain limitations, including prohibitive costs with the need for a large number of evaluations, an inflexibility in defining the search space by often having to select from a limited set of possible design components, and a difficulty of integrating existing architecture code by requiring a specialized design language for search space specification. We propose a simplified search tool, with efficiency in the number of evaluations needed to achieve good results, and flexibility by design, allowing for an easy and open definition of the search space and objective function. Interoperability with existing code or newly released architectures from the literature allows the user to quickly and easily tune architectures to produce well-performing solutions tailor-made for particular use cases. We practically apply this tool to certain vision search spaces, and showcase its effectiveness.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Neural Architecture Search (NAS) consists of applying an optimization technique to find the best performing architecture(s) in a defined search space, with regard to an objective function. The practical implementation of NAS currently carries certain limitations, including prohibitive costs with the need for a large number of evaluations, an inflexibility in defining the search space by often having to select from a limited set of possible design components, and a difficulty of integrating existing architecture code by requiring a specialized design language for search space specification. We propose a simplified search tool, with efficiency in the number of evaluations needed to achieve good results, and flexibility by design, allowing for an easy and open definition of the search space and objective function. Interoperability with existing code or newly released architectures from the literature allows the user to quickly and easily tune architectures to produce well-performing solutions tailor-made for particular use cases. We practically apply this tool to certain vision search spaces, and showcase its effectiveness.
4316.
Xiong, Jianghui; Gu, Shangfeng; Rao, Yuan; Zhang, Xiaodan; Wu, Yuting; Lu, Jie; Jin, Xiu
An innovative fusion method with micro-vision and spectrum of wheat for detecting asymptomatic Fusarium head blight Journal Article
In: Journal of Food Composition and Analysis, vol. 140, pp. 107258, 2025, ISSN: 0889-1575.
@article{XIONG2025107258,
title = {An innovative fusion method with micro-vision and spectrum of wheat for detecting asymptomatic Fusarium head blight},
author = {Jianghui Xiong and Shangfeng Gu and Yuan Rao and Xiaodan Zhang and Yuting Wu and Jie Lu and Xiu Jin},
url = {https://www.sciencedirect.com/science/article/pii/S0889157525000729},
doi = {https://doi.org/10.1016/j.jfca.2025.107258},
issn = {0889-1575},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
journal = {Journal of Food Composition and Analysis},
volume = {140},
pages = {107258},
abstract = {Fusarium head blight (FHB) poses a significant threat to global wheat health and seriously affects the quality of the wheat and its products. Therefore, detection of early FHB infection in wheat is crucial for preventing its rapid spread and ensuring food safety. This study proposed an innovative fusion method for detecting the severity of FHB invasion in wheat based on near-infrared spectroscopy and microscopic visual images. This method concatenated 512 features from near-infrared spectra and microscopic visual images of wheat and used neural architecture search (NAS) to build a model for fused features to achieve accurate classification of the degree of infection caused by pathogens in wheat, with accuracy of 90.60 % and F1-score of 90.95 %. This represented significant improvements of 20.80 % and 21.79 % over single spectral data modelling and 11.41 % and 12.67 % over single image data modelling, respectively. The study results showed that this method enables more accurate and non-destructive detection of FHB in wheat, providing a solution for the early identification of potential fungal diseases, which is valuable for improving the quality and yield of wheat.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Fusarium head blight (FHB) poses a significant threat to global wheat health and seriously affects the quality of the wheat and its products. Therefore, detection of early FHB infection in wheat is crucial for preventing its rapid spread and ensuring food safety. This study proposed an innovative fusion method for detecting the severity of FHB invasion in wheat based on near-infrared spectroscopy and microscopic visual images. This method concatenated 512 features from near-infrared spectra and microscopic visual images of wheat and used neural architecture search (NAS) to build a model for fused features to achieve accurate classification of the degree of infection caused by pathogens in wheat, with accuracy of 90.60 % and F1-score of 90.95 %. This represented significant improvements of 20.80 % and 21.79 % over single spectral data modelling and 11.41 % and 12.67 % over single image data modelling, respectively. The study results showed that this method enables more accurate and non-destructive detection of FHB in wheat, providing a solution for the early identification of potential fungal diseases, which is valuable for improving the quality and yield of wheat.
4315.
Wang, Yifan; Zhong, Linlin
NAS-PINNv2: Improved neural architecture search framework for physics-informed neural networks in low-temperature plasma simulation Technical Report
2025.
@techreport{wang2025naspinnv2improvedneuralarchitecture,
title = {NAS-PINNv2: Improved neural architecture search framework for physics-informed neural networks in low-temperature plasma simulation},
author = {Yifan Wang and Linlin Zhong},
url = {https://arxiv.org/abs/2501.15160},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
keywords = {},
pubstate = {published},
tppubtype = {techreport}
}