Visual Attention Network

Visual Attention Network

摘要

While originally designed for natural language processing (NLP) tasks, the self-attention mechanism has recently taken various computer vision areas by storm. However, the 2D nature of images brings three challenges for applying self-attention in computer vision. (1) Treating images as 1D sequences neglects their 2D structures. (2) The quadratic complexity is too expensive for high-resolution images. (3) It only captures spatial adaptability but ignores channel adaptability. In this paper, we propose a novel large kernel attention (LKA) module to enable self-adaptive and long-range correlations in self-attention while avoiding the above issues. We further introduce a novel neural network based on LKA, namely Visual Attention Network (VAN). While extremely simple and efficient, VAN outperforms the state-of-the-art vision transformers and convolutional neural networks with a large margin in extensive experiments, including image classification, object detection, semantic segmentation, instance segmentation, etc.

结果和模型

ImageNet-1k

模型	预训练	分辨率	参数量（M）	Flops(G)	Top-1 (%)	Top-5 (%)	配置文件	下载
VAN-T*	从头训练	224x224	4.11	0.88	75.41	93.02	config	model
VAN-S*	从头训练	224x224	13.86	2.52	81.01	95.63	config	model
VAN-B*	从头训练	224x224	26.58	5.03	82.80	96.21	config	model
VAN-L*	从头训练	224x224	44.77	8.99	83.86	96.73	config	model

*Models with * are converted from the official repo. The config files of these models are only for validation. We don’t ensure these config files’ training accuracy and welcome you to contribute your reproduction results.

引用

@article{guo2022visual,
  title={Visual Attention Network},
  author={Guo, Meng-Hao and Lu, Cheng-Ze and Liu, Zheng-Ning and Cheng, Ming-Ming and Hu, Shi-Min},
  journal={arXiv preprint arXiv:2202.09741},
  year={2022}
}