Ripple Attention for Visual Perception with Sub-quadratic Complexity

Lin Zheng, Huijie Pan, Lingpeng Kong
Proceedings of the 39th International Conference on Machine Learning, PMLR 162:26993-27010, 2022.

Abstract

Transformer architectures are now central to sequence modeling tasks. At its heart is the attention mechanism, which enables effective modeling of long-term dependencies in a sequence. Recently, transformers have been successfully applied in the computer vision domain, where 2D images are first segmented into patches and then treated as 1D sequences. Such linearization, however, impairs the notion of spatial locality in images, which bears important visual clues. To bridge the gap, we propose ripple attention, a sub-quadratic attention mechanism for vision transformers. Built upon the recent kernel-based efficient attention mechanisms, we design a novel dynamic programming algorithm that weights contributions of different tokens to a query with respect to their relative spatial distances in the 2D space in linear observed time. Extensive experiments and analyses demonstrate the effectiveness of ripple attention on various visual tasks.

Cite this Paper

BibTeX
@InProceedings{pmlr-v162-zheng22a, title = {Ripple Attention for Visual Perception with Sub-quadratic Complexity}, author = {Zheng, Lin and Pan, Huijie and Kong, Lingpeng}, booktitle = {Proceedings of the 39th International Conference on Machine Learning}, pages = {26993--27010}, year = {2022}, editor = {Chaudhuri, Kamalika and Jegelka, Stefanie and Song, Le and Szepesvari, Csaba and Niu, Gang and Sabato, Sivan}, volume = {162}, series = {Proceedings of Machine Learning Research}, month = {17--23 Jul}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v162/zheng22a/zheng22a.pdf}, url = {https://proceedings.mlr.press/v162/zheng22a.html}, abstract = {Transformer architectures are now central to sequence modeling tasks. At its heart is the attention mechanism, which enables effective modeling of long-term dependencies in a sequence. Recently, transformers have been successfully applied in the computer vision domain, where 2D images are first segmented into patches and then treated as 1D sequences. Such linearization, however, impairs the notion of spatial locality in images, which bears important visual clues. To bridge the gap, we propose ripple attention, a sub-quadratic attention mechanism for vision transformers. Built upon the recent kernel-based efficient attention mechanisms, we design a novel dynamic programming algorithm that weights contributions of different tokens to a query with respect to their relative spatial distances in the 2D space in linear observed time. Extensive experiments and analyses demonstrate the effectiveness of ripple attention on various visual tasks.} }
Endnote
%0 Conference Paper %T Ripple Attention for Visual Perception with Sub-quadratic Complexity %A Lin Zheng %A Huijie Pan %A Lingpeng Kong %B Proceedings of the 39th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2022 %E Kamalika Chaudhuri %E Stefanie Jegelka %E Le Song %E Csaba Szepesvari %E Gang Niu %E Sivan Sabato %F pmlr-v162-zheng22a %I PMLR %P 26993--27010 %U https://proceedings.mlr.press/v162/zheng22a.html %V 162 %X Transformer architectures are now central to sequence modeling tasks. At its heart is the attention mechanism, which enables effective modeling of long-term dependencies in a sequence. Recently, transformers have been successfully applied in the computer vision domain, where 2D images are first segmented into patches and then treated as 1D sequences. Such linearization, however, impairs the notion of spatial locality in images, which bears important visual clues. To bridge the gap, we propose ripple attention, a sub-quadratic attention mechanism for vision transformers. Built upon the recent kernel-based efficient attention mechanisms, we design a novel dynamic programming algorithm that weights contributions of different tokens to a query with respect to their relative spatial distances in the 2D space in linear observed time. Extensive experiments and analyses demonstrate the effectiveness of ripple attention on various visual tasks.
APA
Zheng, L., Pan, H. & Kong, L.. (2022). Ripple Attention for Visual Perception with Sub-quadratic Complexity. Proceedings of the 39th International Conference on Machine Learning, in Proceedings of Machine Learning Research 162:26993-27010 Available from https://proceedings.mlr.press/v162/zheng22a.html.

Related Material