AutoML | Literature on Neural Architecture Search

1069.

Xu, Jin; Chen, Mingjian; Huang, Jianqiang; Tang, Xingyuan; Hu, Ke; Li, Jian; Cheng, Jia; Lei, Jun

AutoHEnsGNN: Winning Solution to AutoGraph Challenge for KDD Cup 2020 Technical Report

2021.

1068.

Zhou, Yuanding; Li, Baopu; Wang, Zhihui; Li, Haojie

Video Action Recognition with Neural Architecture Search Proceedings Article

In: Balasubramanian, Vineeth N.; Tsang, Ivor W. (Ed.): Asian Conference on Machine Learning, ACML 2021, 17-19 November 2021, Virtual Event, pp. 1675–1690, PMLR, 2021.

Links | BibTeX

1067.

Gala, Jay; Xie, Pengtao

Learning from Mistakes based on Class Weighting with Application to Neural Architecture Search Technical Report

2021.

Links | BibTeX

1066.

Zhu, Yichen; Du, Jie; Zhu, Yuqin; Wang, Yi; Ou, Zhicai; Feng, Feifei; Tang, Jian

Training BatchNorm Only in Neural Architecture Search and Beyond Technical Report

2021.

Links | BibTeX

1065.

Yang, Longxing; Hu, Yu; Lu, Shun; Sun, Zihao; Mei, Jilin; Zeng, Yiming; Shi, Zhiping; Han, Yinhe; Li, Xiaowei

DDSAS: Dynamic and Differentiable Space-Architecture Search Proceedings Article

In: Balasubramanian, Vineeth N.; Tsang, Ivor W. (Ed.): Asian Conference on Machine Learning, ACML 2021, 17-19 November 2021, Virtual Event, pp. 284–299, PMLR, 2021.

Links | BibTeX

1064.

Zhang, Ruisi; Liang, Youwei; Somayajula, Sai Ashish; Xie, Pengtao

Improving Differentiable Architecture Search with a Generative Model Technical Report

2021.

Links | BibTeX

1063.

Liu, Zechun; Shen, Zhiqiang; Long, Yun; Xing, Eric P.; Cheng, Kwang-Ting; Leichner, Chas

Data-Free Neural Architecture Search via Recursive Label Calibration Technical Report

2021.

Links | BibTeX

1062.

Peng, Cheng; Li, Yangyang; Shang, Ronghua; Jiao, Licheng

RSBNet: One-Shot Neural Architecture Search for A Backbone Network in Remote Sensing Image Recognition Technical Report

2021.

Links | BibTeX

1061.

Do, Tu; Luong, Ngoc Hoang

Training-Free Multi-objective Evolutionary Neural Architecture Search via Neural Tangent Kernel and Number of Linear Regions Proceedings Article

In: Mantoro, Teddy; Lee, Minho; Ayu, Media Anugerah; Wong, Kok Wai; Hidayanto, Achmad Nizar (Ed.): Neural Information Processing - 28th International Conference, ICONIP 2021, Sanur, Bali, Indonesia, December 8-12, 2021, Proceedings, Part II, pp. 335–347, Springer, 2021.

Links | BibTeX

1060.

Liu, Zitu; Zhu, Kai; Liu, Yue; Liu, Qun; Wang, Guoyin; Guo, Yike

Interpretability Analysis Method Based on Path Decomposition for Cell-based Neural Architecture Search Proceedings Article

In: 2021 7th International Conference on Big Data and Information Analytics (BigDIA), pp. 9-16, 2021.

Links | BibTeX

1059.

Cai, Jinhang; Ou, Yimin; Li, Xiu; Wang, Haoqian

ST-NAS: Efficient Optimization of Joint Neural Architecture and Hyperparameter Proceedings Article

In: Mantoro, Teddy; Lee, Minho; Ayu, Media Anugerah; Wong, Kok Wai; Hidayanto, Achmad Nizar (Ed.): Neural Information Processing, pp. 274–281, Springer International Publishing, Cham, 2021, ISBN: 978-3-030-92307-5.

Abstract | BibTeX

1058.

Shi, Jiachen; Fan, Yi; Zhou, Guoqiang; Shen, Jun

Distributed GAN: Towards A Faster Reinforcement Learning Based Architecture Search Journal Article

In: IEEE Transactions on Artificial Intelligence, pp. 1-1, 2021.

Links | BibTeX

1057.

Gong, Saijun; Chen, Lin; Dong, Zhicheng

Neural Network Pruning via Genetic Wavelet Channel Search Proceedings Article

In: Mantoro, Teddy; Lee, Minho; Ayu, Media Anugerah; Wong, Kok Wai; Hidayanto, Achmad Nizar (Ed.): Neural Information Processing, pp. 348–358, Springer International Publishing, Cham, 2021, ISBN: 978-3-030-92270-2.

Abstract | Links | BibTeX

@inproceedings{10.1007/978-3-030-92270-2_30,

title = {Neural Network Pruning via Genetic Wavelet Channel Search},

author = {Saijun Gong and Lin Chen and Zhicheng Dong},

editor = {Teddy Mantoro and Minho Lee and Media Anugerah Ayu and Kok Wai Wong and Achmad Nizar Hidayanto},

url = {https://link.springer.com/chapter/10.1007/978-3-030-92270-2_30},

isbn = {978-3-030-92270-2},

year  = {2021},

date = {2021-01-01},

urldate = {2021-01-01},

booktitle = {Neural Information Processing},

pages = {348--358},

publisher = {Springer International Publishing},

address = {Cham},

abstract = {Neural network pruning has been commonly adopted for alleviating the computational cost of resource-limited devices. Neural architecture search (NAS) is an efficient approach to facilitate neural network compression. However, most existing NAS-based algorithms focus on searching the best sub-network architecture at the layer level, while ignoring the channels that contain richer information. Meanwhile, fixing one pruning rate could not be suitable for all the layers with various features. In this paper, we present a novel NAS-based network pruning strategy called genetic wavelet channel search (GWCS) model, which automatically prunes the network at the channel level with two-fold ideas: (a) Each channel in the whole network is gene coded and pruned adaptively in multiple stages with dynamic genetic selection, and (b) the fitness function of channels is carefully designed by calculating the similarity between the pre-trained network and pruned network using wavelet decomposition to discover the most representative and discriminative channels. Extensive experiments are conducted on CIFAR-10 and CIFAR-100 using ResNet series networks. The proposed GWCS method outperforms other existing pruning approaches in terms of accuracy with a higher compression rate. Our source code is freely available at https://github.com/MGongsj/Network-pruning.git.},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

Close

1056.

Tian, Yunjie; Xie, Lingxi; Fang, Jiemin; Jiao, Jianbin; Ye, Qixiang; Tian, Qi

Exploring Complicated Search Spaces with Interleaving-Free Sampling Technical Report

2021.

Links | BibTeX

1055.

Liu, Jinfeng; Kong, Lingtong; Yang, Jie

Designing and Searching for Lightweight Monocular Depth Network Proceedings Article

In: Mantoro, Teddy; Lee, Minho; Ayu, Media Anugerah; Wong, Kok Wai; Hidayanto, Achmad Nizar (Ed.): Neural Information Processing, pp. 477–488, Springer International Publishing, Cham, 2021, ISBN: 978-3-030-92273-3.

Abstract | Links | BibTeX

@inproceedings{10.1007/978-3-030-92273-3_39,

title = {Designing and Searching for Lightweight Monocular Depth Network},

author = {Jinfeng Liu and Lingtong Kong and Jie Yang},

editor = {Teddy Mantoro and Minho Lee and Media Anugerah Ayu and Kok Wai Wong and Achmad Nizar Hidayanto},

url = {https://link.springer.com/chapter/10.1007/978-3-030-92273-3_39},

isbn = {978-3-030-92273-3},

year  = {2021},

date = {2021-01-01},

urldate = {2021-01-01},

booktitle = {Neural Information Processing},

pages = {477--488},

publisher = {Springer International Publishing},

address = {Cham},

abstract = {Depth sensing is extremely notable in some tasks of robot and autonomous driving. Nowadays, monocular depth estimation based on deep learning becomes a research focus in computer vision. However, most of the current work is seeking for more complex models to get higher accuracy, which can not achieve real-time inference on mobile or embedded systems. Therefore, we aim to design a lightweight model in this paper. At first, we improve the state-of-the-art model, FastDepth, producing FastDepthV2, which has higher accuracy and lower latency on the NYU Depth v2 dataset. Besides, since designing artificial networks takes time and effort, we make it automatic to design lightweight models in monocular depth estimation, using neural architecture search (NAS). Further, inspired by the architecture of MobileNetV2, a factorized hierarchical search space for encoder is used in this paper. In the meanwhile, we incorporate the accuracy and multiply-add operations of a model together into the searching objective, and use gradient-based reinforcement learning algorithm in the searching iterations. The controller in the reinforcement learning framework has converged after more than 1000 searching iterations, and three network architectures with the best performance are obtained. Under the same conditions of training and testing, two of them perform better than FastDepthV2.},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}

Close

1054.

Zhou, Yuguang; He, Zheng; Wan, Tao; Qin, Zengchang

Random Neural Graph Generation with Structure Evolution Proceedings Article

In: Mantoro, Teddy; Lee, Minho; Ayu, Media Anugerah; Wong, Kok Wai; Hidayanto, Achmad Nizar (Ed.): Neural Information Processing, pp. 87–98, Springer International Publishing, Cham, 2021, ISBN: 978-3-030-92270-2.

Abstract | Links | BibTeX

1053.

Laube, Kevin Alexander

The UniNAS framework: combining modules in arbitrarily complex configurations with argument trees Technical Report

2021.

Links | BibTeX

1052.

Nguyen, Giang; Islam, Johir; Pan, Rangeet; Rajan, Hridesh

Manas: Mining Software Repositories to Assist AutoML Technical Report

2021.

Links | BibTeX

1051.

Jiang, Yifan; Gong, Xinyu; Wu, Junru; Shi, Humphrey; Yan, Zhicheng; Wang, Zhangyang

Auto-X3D: Ultra-Efficient Video Understanding via Finer-Grained Neural Architecture Search Technical Report

2021.

Links | BibTeX

1050.

Lu, Shun; Hu, Yu; Yang, Longxing; Sun, Zihao; Mei, Jilin; Yiming, Zeng; Li, Xiaowei

DU-DARTS: Decreasing the Uncertainty of Differentiable Architecture Search Proceedings Article

In: BMVC, 2021.

Links | BibTeX

1049.

Bai, Jinyu; Fan, Yunqian; Sun, Sifan; Kang, Wang; Zhao, Weisheng

Tiny neural network search and implementation for embedded FPGA: a software-hardware co-design approach Proceedings Article

In: 2021 IEEE Asian Solid-State Circuits Conference (A-SSCC), pp. 1-3, 2021.

Links | BibTeX

1048.

Hou, Boyu; Dong, Junwei; Feng, Liang; Qiu, Minghui

Efficient Two-Stage Evolutionary Search of Convolutional Neural Architectures Based on Cell Independence Analysis Proceedings Article

In: Mantoro, Teddy; Lee, Minho; Ayu, Media Anugerah; Wong, Kok Wai; Hidayanto, Achmad Nizar (Ed.): Neural Information Processing, pp. 599–607, Springer International Publishing, Cham, 2021, ISBN: 978-3-030-92307-5.

Abstract | Links | BibTeX

1047.

Liu, Risheng; Ma, Long; Ma, Tengyu; Fan, Xin; Luo, Zhongxuan

Learning with Nested Scene Modeling and Cooperative Architecture Search for Low-Light Vision Technical Report

2021.

Links | BibTeX

1046.

Wang, He; Zhu, Xinshan; Chen, Pinyin; Yang, Yuxuan; Ma, Chao; Gao, Zhongke

A gradient-based automatic optimization CNN framework for EEG state recognition Journal Article

In: Journal of Neural Engineering, 2021.

Abstract | Links | BibTeX

@article{10.1088/1741-2552/ac41ac,

title = {A gradient-based automatic optimization CNN framework for EEG state recognition},

author = {He Wang and Xinshan Zhu and Pinyin Chen and Yuxuan Yang and Chao Ma and Zhongke Gao},

url = {http://iopscience.iop.org/article/10.1088/1741-2552/ac41ac},

year  = {2021},

date = {2021-01-01},

urldate = {2021-01-01},

journal = {Journal of Neural Engineering},

abstract = {The Electroencephalogram (EEG) signal, as a data carrier that can contain a large amount of information about the human brain in different states, is one of the most widely used metrics for assessing human psychophysiological states. Among a variety of analysis methods, deep learning, especially convolutional neural network (CNN), has achieved remarkable results in recent years as a method to effectively extract features from EEG signals. Although deep learning has the advantages of automatic feature extraction and effective classification, it also faces difficulties in network structure design and requires an army of prior knowledge. Automating the design of these hyperparameters can therefore save experts' time and manpower. Neural architecture search techniques have thus emerged. In this paper, based on an existing gradient-based NAS algorithm, PC-DARTS, with targeted improvements and optimizations for the characteristics of EEG signals. Specifically, we establish the model architecture step by step based on the manually designed deep learning models for EEG discrimination by retaining the framework of the search algorithm and performing targeted optimization of the model search space. Corresponding features are extracted separately according to the frequency domain, time domain characteristics of the EEG signal and the spatial position of the EEG electrode. The architecture was applied to EEG-based emotion recognition and driver drowsiness assessment tasks. The results illustrate that compared with the existing methods, the model architecture obtained in this paper can achieve competitive overall accuracy and better standard deviation in both tasks. Therefore, this approach is an effective migration of NAS technology into the field of EEG analysis and has great potential to provide high-performance results for other types of classification and prediction tasks. This can effectively reduce the time cost for researchers and facilitate the application of CNN in more areas.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

Close

The Electroencephalogram (EEG) signal, as a data carrier that can contain a large amount of information about the human brain in different states, is one of the most widely used metrics for assessing human psychophysiological states. Among a variety of analysis methods, deep learning, especially convolutional neural network (CNN), has achieved remarkable results in recent years as a method to effectively extract features from EEG signals. Although deep learning has the advantages of automatic feature extraction and effective classification, it also faces difficulties in network structure design and requires an army of prior knowledge. Automating the design of these hyperparameters can therefore save experts' time and manpower. Neural architecture search techniques have thus emerged. In this paper, based on an existing gradient-based NAS algorithm, PC-DARTS, with targeted improvements and optimizations for the characteristics of EEG signals. Specifically, we establish the model architecture step by step based on the manually designed deep learning models for EEG discrimination by retaining the framework of the search algorithm and performing targeted optimization of the model search space. Corresponding features are extracted separately according to the frequency domain, time domain characteristics of the EEG signal and the spatial position of the EEG electrode. The architecture was applied to EEG-based emotion recognition and driver drowsiness assessment tasks. The results illustrate that compared with the existing methods, the model architecture obtained in this paper can achieve competitive overall accuracy and better standard deviation in both tasks. Therefore, this approach is an effective migration of NAS technology into the field of EEG analysis and has great potential to provide high-performance results for other types of classification and prediction tasks. This can effectively reduce the time cost for researchers and facilitate the application of CNN in more areas.

Close

1045.

Lee, Jaeseong; Rhim, Jungsub; Kang, Duseok; Ha, Soonhoi

S3NAS: Fast Hardware-aware Neural Architecture Search Methodology Journal Article

In: IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, pp. 1-1, 2021.

Links | BibTeX

1044.

Zhou, Yao; Yuan, Xianglei; Zhang, Xiaozhi; Liu, Wei; Wu, Yu; Yen, Gary G.; Hu, Bing; Zhang, Yi

Evolutionary Neural Architecture Search for Automatic Esophageal Lesion Identification and Segmentation Journal Article

In: IEEE Transactions on Artificial Intelligence, pp. 1-1, 2021.

Links | BibTeX

1043.

Chen, Donghui; Chen, Ling; Shang, Zongjiang; Zhang, Youdong; Wen, Bo; Yang, Chenghu

Scale-Aware Neural Architecture Search for Multivariate Time Series Forecasting Technical Report

2021.

Links | BibTeX

1042.

Li, Xiaobin; Jiang, Hongxu; Tian, Fangzheng; Zhang, Yonghua; Miao, Rui; Xu, Donghuan

Optimal FPGA-oriented lightweight network architecture search under multi-objective constraints Proceedings Article

In: 2021 IEEE Intl Conf on Parallel & Distributed Processing with Applications, Big Data & Cloud Computing, Sustainable Computing & Communications, Social Computing & Networking (ISPA/BDCloud/SocialCom/SustainCom), New York City, NY, USA, September 30 - Oct. 3, 2021, pp. 978–983, IEEE, 2021.

Links | BibTeX

1041.

Huang, Zhenhan; Jiang, Chunheng; Chen, Pin-Yu; Gao, Jianxi

Network Graph Based Neural Architecture Search Technical Report

2021.

Links | BibTeX

1040.

Carmichael, Zachariah; Moon, Tim; Jacobs, Sam Ade

Learning Interpretable Models Through Multi-Objective Neural Architecture Search Technical Report

2021.

Links | BibTeX

1039.

Kuang, Huiyu

M-FasterSeg: An Efficient Semantic Segmentation Network Based on Neural Architecture Search Technical Report

2021.

Links | BibTeX

1038.

Kung, Chi-Hsi; Lee, Che-Rung

ADD: A Fine-grained Dynamic Inference Architecture for Semantic Image Segmentation Proceedings Article

In: 2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pp. 4792-4799, 2021.

Links | BibTeX

1037.

Peng, Cheng; Myronenko, Andriy; Hatamizadeh, Ali; Nath, Vishwesh; Siddiquee, Md Mahfuzur Rahman; He, Yufan; Xu, Daguang; Chellappa, Rama; Yang, Dong

HyperSegNAS: Bridging One-Shot Neural Architecture Search with 3D Medical Image Segmentation using HyperNet Technical Report

2021.

Links | BibTeX

1036.

Dong, Xuanyi; Kedziora, David Jacob; Musial, Katarzyna; Gabrys, Bogdan

Automated Deep Learning: Neural Architecture Search Is Not the End Technical Report

2021.

Links | BibTeX

1035.

Shi, Xian; Zhou, Pan; Chen, Wei; Xie, Lei

Efficient Gradient-Based Neural Architecture Search For End-to-End ASR Proceedings Article

In: Companion Publication of the 2021 International Conference on Multimodal Interaction, pp. 91–96, Association for Computing Machinery, Montreal, QC, Canada, 2021, ISBN: 9781450384711.

Abstract | Links | BibTeX