Revealing the Dark Secrets of Extremely Large Kernel ConvNets on Robustness

Honghao Chen, Yurong Zhang, Xiaokun Feng, Xiangxiang Chu, Kaiqi Huang
Proceedings of the 41st International Conference on Machine Learning, PMLR 235:7687-7699, 2024.

Abstract

Robustness is a vital aspect to consider when deploying deep learning models into the wild. Numerous studies have been dedicated to the study of the robustness of vision transformers (ViTs), which have dominated as the mainstream backbone choice for vision tasks since the dawn of 2020s. Recently, some large kernel convnets make a comeback with impressive performance and efficiency. However, it still remains unclear whether large kernel networks are robust and the attribution of their robustness. In this paper, we first conduct a comprehensive evaluation of large kernel convnets’ robustness and their differences from typical small kernel counterparts and ViTs on six diverse robustness benchmark datasets. Then to analyze the underlying factors behind their strong robustness, we design experiments from both quantitative and qualitative perspectives to reveal large kernel convnets’ intriguing properties that are completely different from typical convnets. Our experiments demonstrate for the first time that pure CNNs can achieve exceptional robustness comparable or even superior to that of ViTs. Our analysis on occlusion invariance, kernel attention patterns and frequency characteristics provide novel insights into the source of robustness. Code available at: https://github.com/Lauch1ng/LKRobust.

Cite this Paper

BibTeX
@InProceedings{pmlr-v235-chen24bb, title = {Revealing the Dark Secrets of Extremely Large Kernel {C}onv{N}ets on Robustness}, author = {Chen, Honghao and Zhang, Yurong and Feng, Xiaokun and Chu, Xiangxiang and Huang, Kaiqi}, booktitle = {Proceedings of the 41st International Conference on Machine Learning}, pages = {7687--7699}, year = {2024}, editor = {Salakhutdinov, Ruslan and Kolter, Zico and Heller, Katherine and Weller, Adrian and Oliver, Nuria and Scarlett, Jonathan and Berkenkamp, Felix}, volume = {235}, series = {Proceedings of Machine Learning Research}, month = {21--27 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v235/main/assets/chen24bb/chen24bb.pdf}, url = {https://proceedings.mlr.press/v235/chen24bb.html}, abstract = {Robustness is a vital aspect to consider when deploying deep learning models into the wild. Numerous studies have been dedicated to the study of the robustness of vision transformers (ViTs), which have dominated as the mainstream backbone choice for vision tasks since the dawn of 2020s. Recently, some large kernel convnets make a comeback with impressive performance and efficiency. However, it still remains unclear whether large kernel networks are robust and the attribution of their robustness. In this paper, we first conduct a comprehensive evaluation of large kernel convnets’ robustness and their differences from typical small kernel counterparts and ViTs on six diverse robustness benchmark datasets. Then to analyze the underlying factors behind their strong robustness, we design experiments from both quantitative and qualitative perspectives to reveal large kernel convnets’ intriguing properties that are completely different from typical convnets. Our experiments demonstrate for the first time that pure CNNs can achieve exceptional robustness comparable or even superior to that of ViTs. Our analysis on occlusion invariance, kernel attention patterns and frequency characteristics provide novel insights into the source of robustness. Code available at: https://github.com/Lauch1ng/LKRobust.} }
Endnote
%0 Conference Paper %T Revealing the Dark Secrets of Extremely Large Kernel ConvNets on Robustness %A Honghao Chen %A Yurong Zhang %A Xiaokun Feng %A Xiangxiang Chu %A Kaiqi Huang %B Proceedings of the 41st International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2024 %E Ruslan Salakhutdinov %E Zico Kolter %E Katherine Heller %E Adrian Weller %E Nuria Oliver %E Jonathan Scarlett %E Felix Berkenkamp %F pmlr-v235-chen24bb %I PMLR %P 7687--7699 %U https://proceedings.mlr.press/v235/chen24bb.html %V 235 %X Robustness is a vital aspect to consider when deploying deep learning models into the wild. Numerous studies have been dedicated to the study of the robustness of vision transformers (ViTs), which have dominated as the mainstream backbone choice for vision tasks since the dawn of 2020s. Recently, some large kernel convnets make a comeback with impressive performance and efficiency. However, it still remains unclear whether large kernel networks are robust and the attribution of their robustness. In this paper, we first conduct a comprehensive evaluation of large kernel convnets’ robustness and their differences from typical small kernel counterparts and ViTs on six diverse robustness benchmark datasets. Then to analyze the underlying factors behind their strong robustness, we design experiments from both quantitative and qualitative perspectives to reveal large kernel convnets’ intriguing properties that are completely different from typical convnets. Our experiments demonstrate for the first time that pure CNNs can achieve exceptional robustness comparable or even superior to that of ViTs. Our analysis on occlusion invariance, kernel attention patterns and frequency characteristics provide novel insights into the source of robustness. Code available at: https://github.com/Lauch1ng/LKRobust.
APA
Chen, H., Zhang, Y., Feng, X., Chu, X. & Huang, K.. (2024). Revealing the Dark Secrets of Extremely Large Kernel ConvNets on Robustness. Proceedings of the 41st International Conference on Machine Learning, in Proceedings of Machine Learning Research 235:7687-7699 Available from https://proceedings.mlr.press/v235/chen24bb.html.

Related Material