Quantized CNN: A Unified Approach to Accelerate and Compress Convolutional Networks
The DOI matches its Crossref record; 20 of the 20 references it cites were found in a registry or an index.
Abstract
We are witnessing an explosive development and widespread application of deep neural networks (DNNs) in various fields. However, DNN models, especially a convolutional neural network (CNN), usually involve massive parameters and are computationally expensive, making them extremely dependent on high-performance hardware. This prohibits their further extensions, e.g., applications on mobile devices. In this paper, we present a quantized CNN, a unified approach to accelerate and compress convolutional networks. Guided by minimizing the approximation error of individual layer's response, both fully connected and convolutional layers are carefully quantized. The inference computation can be effectively carried out on the quantized network, with much lower memory and storage consumption. Quantitative evaluation on two publicly available benchmarks demonstrates the promising performance of our approach: with comparable classification accuracy, it achieves 4 to $6 \times $ acceleration and 15 to $20\times $ compression. With our method, accurate image classification can even be directly carried out on mobile devices within 1 s.
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
Gradient-based learning applied to document recognition
Lecun, Y., Bottou, L., Bengio, Y., & Haffner, P. (1998) · Proceedings of the IEEE · doi:10.1109/5.726791 · Cited by 58,805
Greedy layer-wise training of deep networks
Bengio, Y., Lamblin, P., Popovici, D., & Larochelle, H. (2007) · Advances in Neural Information Processing Systems 19 · doi:10.7551/mitpress/7503.003.0024 · Cited by 4,841 · cited 1×
“, sparse autoencoder [37] and sparse deep belief network [38].”
Han, S., Mao, H., & Dally, W. J. (2015) · arXiv · doi:10.48550/arxiv.1510.00149 · Cited by 4,662
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
Return of the Devil in the Details: Delving Deep into Convolutional Nets
Chatfield, K., Simonyan, K., Vedaldi, A., & Zisserman, A. (2014) · Proceedings of the British Machine Vision Conference 2014 · doi:10.5244/c.28.6 · Cited by 3,335
Product Quantization for Nearest Neighbor Search
Jégou, H., Douze, M., & Schmid, C. (2011) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2010.57 · Cited by 2,956 · cited 1×
“Product quantization [52] is a commonly used technique in approximate nearest neighbor search, which can usually offer much higher retrieval accuracy than approaches based on hashing [53], [54].”
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
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
Courbariaux, M., Hubara, I., Soudry, D., El-Yaniv, R., & Bengio, Y. (2016) · arXiv · doi:10.48550/arxiv.1602.02830 · Cited by 2,106
Buciluǎ, C., Caruana, R., & Niculescu-Mizil, A. (2006) · Proceedings of the 12th ACM SIGKDD international conference on Knowledge discovery and data mining · doi:10.1145/1150402.1150464 · Cited by 2,068 · cited 1×
“Ba and Caruana [28] extended the idea of model compression in [51], and demonstrated that the shallow network could be more accurately trained by mimicking the output of a deep network (or an ensemble of deep networks).”
Learning both Weights and Connections for Efficient Neural Networks
Han, S., Pool, J., Tran, J., & Dally, W. J. (2015) · arXiv · doi:10.48550/arxiv.1506.02626 · Cited by 1,481 · cited 1×
“[42] proposed a three-stage pruning method: first identify important connections in the network, then remove unimportant connection, and finally retrain the network with those remaining connections.”
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 6×
“[10] proposed to approximate the weighting matrix as the multiplication of two low-rank matrices, to reduce the parameter redundancy.”
InfluentialDeep Learning with Limited Numerical Precision
Gupta, S., Agrawal, A., Gopalakrishnan, K., & Narayanan, P. (2015) · arXiv · doi:10.48550/arxiv.1502.02551 · Cited by 1,364 · cited 2×
“Other approaches included fixed-point representation [18]–[20], binary-valued network [21]–[24], codebook-based quantization [25]–[27], and compact network design [28]–[31].”
Speeding 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
Compressing Neural Networks with the Hashing Trick
Chen, W., Wilson, J. T., Tyree, S., Weinberger, K. Q., & Chen, Y. (2015) · arXiv · doi:10.48550/arxiv.1504.04788 · Cited by 974
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 5×
“Several methods have been proposed to accelerate and compress CNN models, and shall be briefly reviewed in this section, categorized by LRD, connection pruning, lowprecision representation, codebook-based quantization, and compact network design.”
InfluentialTraining deep neural networks with low precision multiplications
Courbariaux, M., Bengio, Y., & David, J.-P. (2014) · arXiv · doi:10.48550/arxiv.1412.7024 · Cited by 407
Yang, 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 2×
“Then, to compare the fully connected layers’ compression, the following baseline methods are considered: 1) RER [39]: random edge removal; 2) LRD [56]: low-rank decomposition; 3) DK [29]: dark knowledge; 4) NN-ES [26]: neural network with equivalent size; 5) HashNet [26]: hashed neural nets; 6) DPP [40]: data-free par…”
Cited by
Model Compression and Hardware Acceleration for Neural Networks: A Comprehensive Survey
Deng, B. L., Li, G., Han, S., Shi, L., & Xie, Y. (2020) · Proceedings of the IEEE · doi:10.1109/jproc.2020.2976475 · Cited by 996
“When quantizing weights, besides minimizing the weight approximation error, the problem can be formulated as to minimize the response error of each layer [189].”
BackgroundA comprehensive survey on model compression and acceleration
Choudhary, T., Mishra, V., Goswami, A., & Sarangapani, J. (2020) · Artificial Intelligence Review · doi:10.1007/s10462-020-09816-7 · Cited by 513
Delving deep into spatial pooling for squeeze-and-excitation networks
Jin, X., Xie, Y., Wei, X.-S., Zhao, B.-R., Chen, Z.-M., & Tan, X. (2022) · Pattern Recognition · doi:10.1016/j.patcog.2021.108159 · Cited by 278
Li, G., Liu, L., Wang, X., Dong, X., Zhao, P., & Feng, X. (2018) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-01418-6_40 · Cited by 80
CoNNa–Hardware accelerator for compressed convolutional neural networks
Struharik, R. J. R., Vukobratović, B. Z., Erdeljan, A. M., & Rakanović, D. M. (2020) · Microprocessors and Microsystems · doi:10.1016/j.micpro.2020.102991 · Cited by 18
Metadata
- Authors
- Cheng, J., Wu, J., Leng, C., Wang, Y., & Hu, Q.
- Venue
- IEEE Transactions on Neural Networks and Learning Systems
- Published
- 2018-10
- Volume
- 29
- Issue
- 10
- Pages
- 4730-4743
- ISSN
- 2162-237X (ISSN-L)
- 2162-2388
- OpenAlex
- W2775811337
- Cited by
- 126
- References
- 20
- FWCI
- 5.62
- 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.