LKCA: large kernel convolutional attention

Chenghao Li; Pengbo Shi; Qingzi Chen; Jirui Liu; Lingyun Zhu

doi:10.1117/12.3031589

18 July 2024 LKCA: large kernel convolutional attention

Chenghao Li, Pengbo Shi, Qingzi Chen, Jirui Liu, Lingyun Zhu

Proceedings Volume 13179, International Conference on Optics and Machine Vision (ICOMV 2024); 131790M (2024) https://doi.org/10.1117/12.3031589
Event: International Conference on Optics and Machine Vision (ICOMV 2024), 2024, Nanchang, China

Abstract

We revisit the relationship between attention mechanisms and large kernel ConvNets in visual transformers and propose a new spatial attention named Large Kernel Convolution Attention (LKCA). It simplifies the attention operation by replacing it with a single large kernel convolution. LKCA combines the advantages of convolutional neural networks and visual transformers, possessing a large receptive field, locality, and parameter sharing. We explained the superiority of LKCA from both convolution and attention perspectives, providing equivalent code implementations for each view. Experiments confirm that LKCA implemented from both the convolutional and attention perspectives exhibit equivalent performance. We conducted experiments with LKCA on a wide range of ViT variants, consistently improving classification and performance compare to various models, showcasing the capabilities of LKCA. Our code will be made publicly available

(2024) Published by SPIE. Downloading of the abstract is permitted for personal use only.

Citation Download Citation

Chenghao Li, Pengbo Shi, Qingzi Chen, Jirui Liu, and Lingyun Zhu "LKCA: large kernel convolutional attention", Proc. SPIE 13179, International Conference on Optics and Machine Vision (ICOMV 2024), 131790M (18 July 2024); https://doi.org/10.1117/12.3031589

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available

Members: $17.00

Non-members: $21.00 ADD TO CART

PROCEEDINGS
6 PAGES

DOWNLOAD PAPER SAVE TO MY LIBRARY

GET CITATION

RIGHTS & PERMISSIONS

Get copyright permission Get copyright permission on Copyright Marketplace

KEYWORDS

Convolution

Transformers

Visualization

Matrices

Visual process modeling

Windows

Modeling

Show All Keywords

Keywords/Phrases

Search In:

Publication Years