Deep Neural Network Compression by In-Parallel Pruning-Quantization
The DOI matches its Crossref record; 32 of the 32 references it cites were found in a registry or an index.
Abstract
Deep neural networks enable state-of-the-art accuracy on visual recognition tasks such as image classification and object detection. However, modern networks contain millions of learned connections, and the current trend is towards deeper and more densely connected architectures. This poses a challenge to the deployment of state-of-the-art networks on resource-constrained systems, such as smartphones or mobile robots. In general, a more efficient utilization of computation resources would assist in deployment scenarios from embedded platforms to computing clusters running ensembles of networks. In this paper, we propose a deep network compression algorithm that performs weight pruning and quantization jointly, and in parallel with fine-tuning. Our approach takes advantage of the complementary nature of pruning and quantization and recovers from premature pruning errors, which is not possible with two-stage approaches. In experiments on ImageNet, CLIP-Q (Compression Learning by In-Parallel Pruning-Quantization) improves the state-of-the-art in network compression on AlexNet, VGGNet, GoogLeNet, and ResNet. We additionally demonstrate that CLIP-Q is complementary to efficient network architecture design by compressing MobileNet and ShuffleNet, and that CLIP-Q generalizes beyond convolutional networks by compressing a memory network for visual question answering.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Deep Residual Learning for Image Recognition
He, K., Zhang, X., Ren, S., & Sun, J. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.90 · Cited by 235,474
ImageNet classification with deep convolutional neural networks
Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2017) · Communications of the ACM · doi:10.1145/3065386 · Cited by 112,155
You Only Look Once: Unified, Real-Time Object Detection
Redmon, J., Divvala, S., Girshick, R., & Farhadi, A. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.91 · Cited by 53,803
Xception: Deep Learning with Depthwise Separable Convolutions
Chollet, F. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.195 · Cited by 20,380
ShuffleNet: An Extremely Efficient Convolutional Neural Network for Mobile Devices
Zhang, X., Zhou, X., Lin, M., & Sun, J. (2017) · arXiv · doi:10.48550/arxiv.1707.01083 · Cited by 9,982
ShuffleNet: An Extremely Efficient Convolutional Neural Network for Mobile Devices
Zhang, X., Zhou, X., Lin, M., & Sun, J. (2018) · 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2018.00716 · Cited by 9,982
Multi-scale context aggregation by dilated convolutions
Yu, F., & Koltun, V. (2015) · arXiv · doi:10.48550/arxiv.1511.07122 · Cited by 7,738 · cited 1×
“DEEP neural networks have become indispensable tools for a wide range of visual recognition tasks, such as image classification [1], [2], [3], object detection [4], [5], [6], semantic segmentation [7], [8], [9], [10], and visual question”
A comparative study of texture measures with classification based on featured distributions
Ojala, T., Pietikäinen, M., & Harwood, D. (1996) · Pattern Recognition · doi:10.1016/0031-3203(95)00067-4 · Cited by 6,882 · cited 1×
“tional layers with a learnable module inspired by local binary patterns [48].”
BackgroundPractical Bayesian Optimization of Machine Learning Algorithms
Snoek, J., Larochelle, H., & Adams, R. P. (2012) · arXiv · doi:10.48550/arxiv.1206.2944 · Cited by 6,691
SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size
Iandola, F. N., Han, S., Moskewicz, M. W., Ashraf, K., Dally, W. J., & Keutzer, K. (2016) · arXiv · doi:10.48550/arxiv.1602.07360 · Cited by 6,201
Gaussian Processes in Machine Learning
Rasmussen, C. E. (2004) · Lecture Notes in Computer Science · doi:10.1007/978-3-540-28650-9_4 · Cited by 5,397
Han, S., Mao, H., & Dally, W. J. (2015) · arXiv · doi:10.48550/arxiv.1510.00149 · Cited by 4,662
VQA: Visual Question Answering
Antol, S., Agrawal, A., Lu, J., Mitchell, M., Batra, D., Zitnick, C. L., & Parikh, D. (2015) · 2015 IEEE International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2015.279 · Cited by 4,487
XNOR-Net: ImageNet Classification Using Binary Convolutional Neural Networks
Rastegari, M., Ordonez, V., Redmon, J., & Farhadi, A. (2016) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-46493-0_32 · Cited by 4,408
Semantic Image Segmentation with Deep Convolutional Nets and Fully Connected CRFs
Chen, L.-C., Papandreou, G., Kokkinos, I., Murphy, K., & Yuille, A. L. (2014) · arXiv · doi:10.48550/arxiv.1412.7062 · Cited by 4,213 · cited 1×
“DEEP neural networks have become indispensable tools for a wide range of visual recognition tasks, such as image classification [1], [2], [3], object detection [4], [5], [6], semantic segmentation [7], [8], [9], [10], and visual question”
FitNets: Hints for Thin Deep Nets
Romero, A., Ballas, N., Kahou, S. E., Chassang, A., Gatta, C., & Bengio, Y. (2014) · arXiv · doi:10.48550/arxiv.1412.6550 · Cited by 2,859
Measurement Reliability and Validity
(2000) · Research Methods in Applied Settings · doi:10.4324/9781410605337-29 · Cited by 2,541
BinaryConnect: Training Deep Neural Networks with binary weights during propagations
Courbariaux, M., Bengio, Y., & David, J.-P. (2015) · arXiv · doi:10.48550/arxiv.1511.00363 · Cited by 2,180
Deep Networks with Stochastic Depth
Huang, G., Sun, Y., Liu, Z., Sedra, D., & Weinberger, K. Q. (2016) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-46493-0_39 · Cited by 2,104
Exploiting Linear Structure Within Convolutional Networks for Efficient Evaluation
Denton, R., Zaremba, W., Bruna, J., LeCun, Y., & Fergus, R. (2014) · arXiv · doi:10.48550/arxiv.1404.0736 · Cited by 1,414 · cited 1×
“Low-rank factorization methods exploit the redundancy in filters and feature map responses [51], [52], [53].”
MethodSpeeding up Convolutional Neural Networks with Low Rank Expansions
Jaderberg, M., Vedaldi, A., & Zisserman, A. (2014) · Proceedings of the British Machine Vision Conference 2014 · doi:10.5244/c.28.88 · Cited by 1,164
Predicting Parameters in Deep Learning
Denil, M., Shakibi, B., Dinh, L., Ranzato, M., & de Freitas, N. (2013) · arXiv · doi:10.48550/arxiv.1306.0543 · Cited by 853 · cited 1×
“has been trained, there will typically be a high degree of parameter redundancy [26].”
BackgroundTo prune, or not to prune: exploring the efficacy of pruning for model compression
Zhu, M., & Gupta, S. (2017) · arXiv · doi:10.48550/arxiv.1710.01878 · Cited by 805 · cited 1×
“On MobileNet, CLIP-Q achieves a substantially higher compression rate compared with [66] and [67], and a 9 percent relative improvement with respect to weighted-entropy quantization [42].”
ResultVisual7W: Grounded Question Answering in Images
Zhu, Y., Groth, O., Bernstein, M., & Fei-Fei, L. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.540 · Cited by 794 · cited 1×
“The task integrates computer vision, natural language processing, and logical reasoning, and is sometimes considered to be “AI-complete” or a proxy to evaluating a system for deep image understanding [68], [69].”
BackgroundDesigning Energy-Efficient Convolutional Neural Networks Using Energy-Aware Pruning
Yang, T.-J., Chen, Y.-H., & Sze, V. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.643 · Cited by 732 · cited 2×
“Smaller network models are not necessarily more energy efficient because energy consumption is also determined by the pattern of memory accesses [40].”
BackgroundMethodImageNet Large Scale Visual Recognition Challenge
Russakovsky, O., Deng, J., Su, H., Krause, J., Satheesh, S., Ma, S., Huang, Z., Karpathy, A., Khosla, A., Bernstein, M., et al. (2014) · arXiv · doi:10.48550/arxiv.1409.0575 · Cited by 567
Predicting Parameters in Deep Learning
Shakibi, B. (2014) · University of British Columbia · doi:10.14288/1.0165555 · Cited by 462
Local Binary Convolutional Neural Networks
Juefei-Xu, F., Boddeti, V. N., & Savvides, M. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.456 · Cited by 297 · cited 1×
“Local binary convolutional networks [47] replace convolu-”
Gysel, P., Pimentel, J., Motamedi, M., & Ghiasi, S. (2018) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2018.2808319 · Cited by 250
CLIP-Q: Deep Network Compression Learning by In-parallel Pruning-Quantization
Tung, F., & Mori, G. (2018) · 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2018.00821 · Cited by 220 · cited 1×
“A preliminary version of this paper appeared in [27].”
BackgroundSelf-citationYang, Z., Moczulski, M., Denil, M., De Freitas, N., Song, L., & Wang, Z. (2015) · 2015 IEEE International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2015.173 · Cited by 215 · cited 1×
“For example, circulant neural networks [49] replace fully connected layers with learned circulant projections, which can be computed quickly using the Fast Fourier Transform, and deep fried convnets [50] replace fully connected layers with an adaptive version of the Fastfood transform.”
MethodNetwork Sketching: Exploiting Binary Structure in Deep CNNs
Guo, Y., Yao, A., Zhao, H., & Chen, Y. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.430 · Cited by 81 · cited 1×
“Network sketching [46] binarizes pre-trained networks by approximating realvalued filters with weighted combinations of binary filters.”
Method
Cited by
Pruning by explaining: A novel criterion for deep neural network pruning
Yeom, S.-K., Seegerer, P., Lapuschkin, S., Binder, A., Wiedemann, S., Müller, K.-R., & Samek, W. (2021) · Pattern Recognition · doi:10.1016/j.patcog.2021.107899 · Cited by 263
Data-Free Network Quantization With Adversarial Knowledge Distillation
Choi, Y., Choi, J., El-Khamy, M., & Lee, J. (2020) · 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw50498.2020.00363 · Cited by 135
“work quantization reduces the memory footprint for weights and activations by quantization and is usually followed by lossless source coding for compression [13–18].”
ChannelNets: Compact and Efficient Convolutional Neural Networks via Channel-Wise Convolutions
Gao, H., Wang, Z., Cai, L., & Ji, S. (2020) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2020.2975796 · Cited by 81
A Generalization Performance Study Using Deep Learning Networks in Embedded Systems
Gorospe, J., Mulero, R., Arbelaitz, O., Muguerza, J., & Antón, M. Á. (2021) · Sensors · doi:10.3390/s21041031 · Cited by 32
“For example, the compression learning by in-parallel pruning-quantization (CLIPQ) [36] method achieves compression ratios up to 72-times lower than the original network, and the accuracy can be improved by up to 0.”
Method
Metadata
- Authors
- Tung, F., & Mori, G.
- Venue
- IEEE Transactions on Pattern Analysis and Machine Intelligence
- Published
- 2020-03-01
- Volume
- 42
- Issue
- 3
- Pages
- 568-579
- ISSN
- 0162-8828 (ISSN-L)
- 2160-9292
- 1939-3539
- OpenAlex
- W2903711666
- Cited by
- 140
- References
- 32
- FWCI
- 6.23
- Type
- journal article
- Language
- en
- Open access
- closed
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.