Student or educator? Sign up with your school email and get 6 months of Pro free.
Cited by 1,477

The DOI matches its Crossref record; 15 of the 15 references it cites were found in a registry or an index.

Dynamic Convolution: Attention Over Convolution Kernels

Open access
Chen, Yinpeng, Dai, Xiyang, Liu, Mengchen, Chen, Dongdong, Yuan, Lu, Liu, Zicheng·2020·2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

Abstract

Light-weight convolutional neural networks (CNNs) suffer performance degradation as their low computational budgets constrain both the depth (number of convolution layers) and the width (number of channels) of CNNs, resulting in limited representation capability. To address this issue, we present Dynamic Convolution, a new design that increases model complexity without increasing the network depth or width. Instead of using a single convolution kernel per layer, dynamic convolution aggregates multiple parallel convolution kernels dynamically based upon their attentions, which are input dependent. Assembling multiple kernels is not only computationally efficient due to the small kernel size, but also has more representation power since these kernels are aggregated in a non-linear way via attention. By simply using dynamic convolution for the state-of-the-art architecture MobileNetV3-Small, the top-1 accuracy of ImageNet classification is boosted by 2.9% with only 4% additional FLOPs and 2.9 AP gain is achieved on COCO keypoint detection.

Overview

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1109/cvpr42600.2020.01104

Cites

15 of 15 references verified0 citation errors · 0 not checkable

Statistics

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

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

Show all 15 references

Research works printed but not linked: 29

  1. hongyi, 0, mixup: Beyond empirical risk minimization, International Conference on Learning Representations
  2. zhang, 0, Lq-nets: Learned quantization for highly accurate and compact deep neural networks, the European Conference on Computer Vision (ECCV)
  3. 10.1007/978-3-030-01231-1_29 · DOI · Paper page
  4. zuxuan, 0, Blockdrop: Dynamic inference paths in residual networks, The IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
  5. wu, 2017, Shift A zero flop zero parameter alternative to spatial convolutions
  6. 10.1109/CVPR.2019.01099 · DOI · Paper page
  7. yu, 0, Slimmable neural networks, International Conference on Learning Representations
  8. 10.1109/CVPR.2019.00748 · DOI · Paper page
  9. brandon, 2019, Condconv: Conditionally parameterized convolutions for efficient inference, NeurIPS
  10. 10.1109/CVPR.2018.00716 · DOI · Paper page
  11. howard, 2017, Mobilenets Efficient convolutional neural networks for mobile vision applications
  12. huang, 0, Multi-scale dense networks for resource efficient image classification, International Conference on Learning Representations
  13. iandola, 2016, Squeezenet: Alexnet-level accuracy with 50x fewer parameters and <1mb model size, CoRR abs/1602 07360
  14. lin, 2017, Runtime neural pruning, Advances in neural information processing systems, 2181
  15. lin, 0, Microsoft coco: Common objects in context, European Conference on Computer Vision, 740
  16. liu, 0, DARTS: Differentiable architecture search, International Conference on Learning Representations
  17. 10.1109/CVPR.2019.00881 · DOI · Paper page
  18. courbariaux, 2015, Binaryconnect: Training deep neural networks with binary weights during propagations, Advances in Neural IInformation Processing Systems, 3123
  19. guo, 2019, Single path one-shot neural architecture search with uniform sampling
  20. 10.1007/978-3-030-01261-8_25 · DOI · Paper page
  21. 10.1109/ICCV.2019.00200 · DOI · Paper page
  22. han, 0, Deep compression: Compressing deep neural networks with pruning, trained quantization and huffman coding, International Conference on Learning Representations (ICLR), 10
  23. cai, 0, ProxylessNAS: Direct neural architecture search on target task and hardware, International Conference on Learning Representations
  24. 10.1007/978-3-030-01234-2_48 · DOI · Paper page
  25. liu, 0, Dynamic deep neural networks: Optimizing accuracy -efficiency trade-offs by selective execution, AAAI Conference on Artificial Intelligence (AAAI)
  26. 10.1609/aaai.v33i01.33014780 · DOI · Paper page
  27. zhu, 0, Trained ternary quantization, International Conference on Learning Representations (ICLR)
  28. nair, 2010, Hinton. Rectified linear units improve restricted boltzmann machines, ICML
  29. 10.1109/CVPR.2018.00907 · DOI · Paper page

Cited by

See all 1,477

Metadata

Authors
Chen, Y., Dai, X., Liu, M., Chen, D., Yuan, L., & Liu, Z.
Venue
2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
Publisher
IEEE
Published
2020-06
Pages
11027-11036
ISSN
1063-6919 (ISSN-L)
OpenAlex
W3034421924
Cited by
1,477
References
15
FWCI
44.36
Type
conference paper
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

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Dynamic Convolution: Attention Over Convolution Kernels. https://app.cochat.ai/papers/10.1109/cvpr42600.2020.01104
The paper
Chen, Y., Dai, X., Liu, M., Chen, D., Yuan, L., & Liu, Z. (2020). Dynamic Convolution: Attention Over Convolution Kernels. 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). https://doi.org/10.1109/cvpr42600.2020.01104

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.