Structured Binary Neural Networks for Image Recognition
Open accessThe DOI matches its Crossref record; 36 of the 36 references it cites were found in a registry or an index.
Abstract
In this paper, we propose to train binarized convolutional neural networks (CNNs) that are of significant importance for deploying deep learning to mobile devices with limited power capacity and computing resources. Previous works on quantizing CNNs often seek to approximate the floating-point information of weights and/or activations using a set of discrete values. Such methods, termed value approximation here, typically are built on the same network architecture of the full-precision counterpart. Instead, we take a new “structured approximation” view for network quantization — it is possible and valuable to exploit flexible architecture transformation when learning low-bit networks, which can achieve even better performance than the original networks in some cases. In particular, we propose a “group decomposition” strategy, termed GroupNet, which divides a network into desired groups. Interestingly, with our GroupNet strategy, each full-precision group can be effectively reconstructed by aggregating a set of homogeneous binary branches. We also propose to learn effective connections among groups to improve the representation capability. To improve the model capacity, we propose to dynamically execute sparse binary branches conditioned on input features while preserving the computational cost. More importantly, the proposed GroupNet shows strong flexibility for a few vision tasks. For instance, we extend the GroupNet for accurate semantic segmentation by embedding the rich context into the binary structure. The proposed GroupNet also shows strong performance on object detection. Experiments on image classification, semantic segmentation, and object detection tasks demonstrate the superior performance of the proposed methods over various quantized networks in the literature. Moreover, the speedup and runtime memory cost evaluation comparing with related quantization strategies is analyzed on GPU platforms, which serves as a strong benchmark for further research.
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
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
Feature Pyramid Networks for Object Detection
Lin, T.-Y., Dollar, P., Girshick, R., He, K., Hariharan, B., & Belongie, S. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.106 · Cited by 30,939
Quantization and Training of Neural Networks for Efficient Integer-Arithmetic-Only Inference
Jacob, B., Kligys, S., Chen, B., Zhu, M., Tang, M., Howard, A., Adam, H., & Kalenichenko, D. (2018) · 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2018.00286 · Cited by 4,114
MnasNet: Platform-Aware Neural Architecture Search for Mobile
Tan, M., Chen, B., Pang, R., Vasudevan, V., Sandler, M., Howard, A., & Le, Q. V. (2019) · 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2019.00293 · Cited by 3,416 · cited 1×
“MNasNet [78] propose to automatically search for mobile CNNs which achieve better mAP quality than MobileNets for COCO object detection.”
MethodSelf-citationBiSeNet: Bilateral Segmentation Network for Real-Time Semantic Segmentation
Yu, C., Wang, J., Peng, C., Gao, C., Yu, G., & Sang, N. (2018) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-01261-8_20 · Cited by 2,968
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
Channel Pruning for Accelerating Very Deep Neural Networks
He, Y., Zhang, X., & Sun, J. (2017) · 2017 IEEE International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2017.155 · Cited by 2,647 · cited 1×
“To solve this, many existing works focus on network pruning [7], [8], [9], low-bit quantization [10], [11] and/or efficient architecture design [12], [13].”
BackgroundBinaryConnect: 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
Estimating or Propagating Gradients Through Stochastic Neurons for Conditional Computation
Bengio, Y., Léonard, N., & Courville, A. (2013) · arXiv · doi:10.48550/arxiv.1308.3432 · Cited by 1,910 · cited 1×
“In general, sign(·) is non-differentiable and so we adopt the straight-through estimator Bengio et al. (2013) (STE) to approximate the gradient calculation.”
MethodDoReFa-Net: Training Low Bitwidth Convolutional Neural Networks with Low Bitwidth Gradients
Zhou, S., Wu, Y., Ni, Z., Zhou, X., Wen, H., & Zou, Y. (2016) · arXiv · doi:10.48550/arxiv.1606.06160 · Cited by 1,690 · cited 7×
“Note that the numerical difference is only a scalar whether activations are represented by {−1, 1} or {0, 1} in bitwise operations to accelerate the dot products [20,22].”
BackgroundMethodInfluentialHubara, I., Soudry, D., & Yaniv, R. E. (2016) · arXiv · doi:10.48550/arxiv.1602.02505 · Cited by 1,667
FINN: A Framework for Fast, Scalable Binarized Neural Network Inference
Umuroglu, Y., Fraser, N. J., Gambardella, G., Blott, M., Leong, P., Jahre, M., & Vissers, K. (2017) · Proceedings of the 2017 ACM/SIGDA International Symposium on Field-Programmable Gate Arrays · doi:10.1145/3020078.3021744 · Cited by 1,189
TVM: an automated end-to-end optimizing compiler for deep learning
(2018) · arXiv (Cornell University) · doi:10.5555/3291168.3291211 · Cited by 1,170
ESPNet: Efficient Spatial Pyramid of Dilated Convolutions for Semantic Segmentation
Mehta, S., Rastegari, M., Caspi, A., Shapiro, L., & Hajishirzi, H. (2018) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-01249-6_34 · Cited by 1,029
Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity
Fedus, W., Zoph, B., & Shazeer, N. (2021) · arXiv · doi:10.48550/arxiv.2101.03961 · Cited by 1,005
HAQ: Hardware-Aware Automated Quantization With Mixed Precision
Wang, K., Liu, Z., Lin, Y., Lin, J., & Han, S. (2019) · 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2019.00881 · Cited by 978
Zhu, C., Han, S., Mao, H., & Dally, W. J. (2016) · arXiv · doi:10.48550/arxiv.1612.01064 · Cited by 865
Efficient Piecewise Training of Deep Structured Models for Semantic Segmentation
Lin, G., Shen, C., van den Hengel, A., & Reid, I. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.348 · Cited by 862
Self-citationLi, F., Liu, B., Wang, X., Zhang, B., & Yan, J. (2016) · arXiv · doi:10.48550/arxiv.1605.04711 · Cited by 732
PACT: Parameterized Clipping Activation for Quantized Neural Networks
Choi, J., Wang, Z., Venkataramani, S., Chuang, P. I.-J., Srinivasan, V., & Gopalakrishnan, K. (2018) · arXiv · doi:10.48550/arxiv.1805.06085 · Cited by 703 · cited 1×
“For the uniform quantizer, works in [40], [41] explicitly parameterize and optimize the upper and/or lower bound of the activation and weights.”
MethodESPNetv2: A Light-Weight, Power Efficient, and General Purpose Convolutional Neural Network
Mehta, S., Rastegari, M., Shapiro, L., & Hajishirzi, H. (2019) · 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2019.00941 · Cited by 551 · cited 1×
“ESPNet [53], [60] design efficient spatial pyramid for real-time semantic segmentation under resource constraints.”
MethodDeep Learning with Low Precision by Half-Wave Gaussian Quantization
Cai, Z., He, X., Sun, J., & Vasconcelos, N. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.574 · Cited by 498 · cited 3×
“The second category approaches focus on improving the quantization function [22], [23], [24], by learning suitable mappings between discrete values and their floating-point counterparts.”
BackgroundMethodTowards Accurate Binary Convolutional Neural Network
Lin, X., Zhao, C., & Pan, W. (2017) · arXiv · doi:10.48550/arxiv.1711.11294 · Cited by 453 · cited 8×
“, multiple binarizations [27], [28], [29], [30], [31]).”
BackgroundMethodInfluentialLearning to Quantize Deep Networks by Optimizing Quantization Intervals With Task Loss
Jung, S., Son, C., Lee, S., Son, J., Han, J.-J., Kwak, Y., Hwang, S. J., & Choi, C. (2019) · 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2019.00448 · Cited by 385 · cited 1×
“For the uniform quantizer, works in [37], [38] explicitly parameterize and optimize the upper and/or lower bound of the activation and weights.”
MethodAI Benchmark: Running Deep Neural Networks on Android Smartphones
Ignatov, A., Timofte, R., Chou, W., Wang, K., Wu, M., Hartley, T., & Van Gool, L. (2019) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-11021-5_19 · Cited by 347
Apprentice: Using Knowledge Distillation Techniques To Improve Low-Precision Network Accuracy
Mishra, A., & Marr, D. (2017) · arXiv · doi:10.48550/arxiv.1711.05852 · Cited by 270
How to Train a Compact Binary Neural Network with High Accuracy?
Tang, W., Hua, G., & Wang, L. (2017) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v31i1.10862 · Cited by 241 · cited 5×
“rforms LBD by 7.2% on the Top-1 accuracy. Note that LBD approach can be treated as a kind of tensor approximation which has similarities with multiple binarizations methods in [43], [44], [45], [46], [47] and the differences are described in Section4. It strongly shows the necessity for employing the group-wise decomp…”
BackgroundMethodInfluentialFully Quantized Network for Object Detection
Li, R., Wang, Y., Liang, F., Qin, H., Yan, J., & Fan, R. (2019) · 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2019.00292 · Cited by 214 · cited 2×
“We compare with the state-of-the-art quantized object detector FQN [26], and also quantize the detector with LQNet using the proposed training strategy.”
BackgroundMethodProgressive Neural Architecture Search
Liu, C., Zoph, B., Neumann, M., Shlens, J., Hua, W., Li, L.-J., Fei-Fei, L., Yuille, A., Huang, J., & Murphy, K. (2017) · arXiv · doi:10.48550/arxiv.1712.00559 · Cited by 171 · cited 2×
“To avoid handcrafted heuristics to explore the enormous design space, NAS [30], [31], [32], [59], [60] has been explored for automatically sample the design space and improve the model quality.”
MethodBinary Ensemble Neural Network: More Bits per Network or More Networks per Bit?
Zhu, S., Dong, X., & Su, H. (2019) · 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2019.00506 · Cited by 153 · cited 1×
“Since we employ binary weights and binary activations, we directly compare to the previous state-of-the-art binary approaches, including BNN [14], XNOR-Net [15], Bi-Real Net [82] and ABC-Net [45] and BENN [94].”
MethodSYQ: Learning Symmetric Quantization for Efficient Deep Neural Networks
Faraone, J., Fraser, N., Blott, M., & Leong, P. H. W. (2018) · 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2018.00452 · Cited by 134
Loss-aware Weight Quantization of Deep Networks
Hou, L., & Kwok, J. T. (2018) · arXiv · doi:10.48550/arxiv.1802.08635 · Cited by 99
Network 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 5×
“[10], [19], [22], [24] QK QK +, −, × ∼ 32 K × < 64 K2 × [43], [44], [45], [46], [47] K ×B K ×B +, −, XNOR-popcount ∼ 32 K × < 64 K2 × Group-Net K × (B,B) +,−, XNOR-popcount ∼ 32 K × < 64 K ×”
BackgroundMethodInfluentialProxQuant: Quantized Neural Networks via Proximal Operators
Bai, Y., Wang, Y.-X., & Liberty, E. (2018) · arXiv · doi:10.48550/arxiv.1810.00861 · Cited by 76 · cited 1×
“A typical approach is to train with regularization [40], [41], where the optimization problem becomes continuous while gradually adjusting the data distribution towards quantization levels.”
BackgroundAQD: Towards Accurate Quantized Object Detection
Chen, P., Liu, J., Zhuang, B., Tan, M., & Shen, C. (2021) · 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr46437.2021.00017 · Cited by 29
Self-citation
Cited by
Rethinking Differentiable Search for Mixed-Precision Neural Networks
Cai, Z., & Vasconcelos, N. (2020) · 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr42600.2020.00242 · Cited by 119
“Uniform low-precision: Low-precision networks have recently become popular to speed-up and reduce model size of deep networks [14, 23, 35, 4, 37, 34, 36].”
BackgroundBiViT: Extremely Compressed Binary Vision Transformers
He, Y., Lou, Z., Zhang, L., Liu, J., Wu, W., Zhou, H., & Zhuang, B. (2023) · 2023 IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv51070.2023.00520 · Cited by 31
“They are mainly composed of 1 × 1 convolutions, which are widely recognized to be difficult to binarize due to the limited representational capability [86, 23, 7].”
BackgroundSelf-citationAQD: Towards Accurate Quantized Object Detection
Chen, P., Liu, J., Zhuang, B., Tan, M., & Shen, C. (2021) · 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr46437.2021.00017 · Cited by 29
“Following [20,41], we use the COCO trainval35k split (115K images) for training and minival split (5K images) for validation.”
Self-citationBoolNet: Minimizing The Energy Consumption of Binary Neural Networks
Guo, N., Bethge, J., Yang, H., Zhong, K., Ning, X., Meinel, C., & Wang, Y. (2021) · arXiv · doi:10.48550/arxiv.2106.06991 · Cited by 10
“…and the need for 32-bit operations. accuracy gap, including specific architecture design [30, 4, 3, 29], real-valued weight and activation approximation [27, 48], specific training recipes [34], a dedicated optimizer [20], leveraging neural architecture search [6, 46] and dynamic networks [7].”
Metadata
- Authors
- Zhuang, B., Shen, C., Tan, M., Chen, P., Liu, L., & Reid, I.
- Venue
- International Journal of Computer Vision
- Published
- 2022-06-22
- Volume
- 130
- Issue
- 9
- Pages
- 2081-2102
- ISSN
- 0920-5691 (ISSN-L)
- 1573-1405
- OpenAlex
- W2974110043
- Cited by
- 22
- References
- 36
- FWCI
- 1.29
- Type
- journal article
- Language
- en
- Open access
- green
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.