Neighborhood Attention Transformer

We present Neighborhood Attention (NA), the first efficient and scalable sliding-window attention mechanism for vision. NA is a pixel-wise operation, localizing self attention (SA) to the nearest neighboring pixels, and therefore enjoys a linear time and space complexity compared to the quadratic complexity of SA. The sliding-window pattern allows NA's receptive field to grow without needing extra pixel shifts, and preserves translational equivariance, unlike Swin Transformer's Window Self Attention (WSA). We develop NATTEN (Neighborhood Attention Extension), a Python package with efficient C++ and CUDA kernels, which allows NA to run up to 40% faster than Swin's WSA while using up to 25% less memory. We further present Neighborhood Attention Transformer (NAT), a new hierarchical transformer design based on NA that boosts image classification and downstream vision performance. Experimental results on NAT are competitive; NAT-Tiny reaches 83.2% top-1 accuracy on ImageNet, 51.4% mAP on MS-COCO and 48.4% mIoU on ADE20K, which is 1.9% ImageNet accuracy, 1.0% COCO mAP, and 2.6% ADE20K mIoU improvement over a Swin model with similar size. To support more research based on sliding-window attention, we open source our project and release our checkpoints at: https://github.com/SHI-Labs/Neighborhood-Attention-Transformer .

PDF Abstract CVPR 2023 PDF CVPR 2023 Abstract

Results from the Paper


Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Semantic Segmentation ADE20K NAT-Base Validation mIoU 49.7 # 126
Params (M) 123 # 30
GFLOPs (512 x 512) 1137 # 19
Semantic Segmentation ADE20K NAT-Small Validation mIoU 49.5 # 131
Params (M) 82 # 39
GFLOPs (512 x 512) 1010 # 16
Semantic Segmentation ADE20K NAT-Tiny Validation mIoU 48.4 # 145
Params (M) 58 # 50
GFLOPs (512 x 512) 934 # 12
Semantic Segmentation ADE20K NAT-Mini Validation mIoU 46.4 # 176
Params (M) 50 # 54
GFLOPs (512 x 512) 900 # 11
Image Classification ImageNet NAT-Mini Top 1 Accuracy 81.8% # 603
Number of params 20M # 585
GFLOPs 2.7 # 171
Image Classification ImageNet NAT-Base Top 1 Accuracy 84.3% # 324
Number of params 90M # 921
GFLOPs 13.7 # 355
Image Classification ImageNet NAT-Tiny Top 1 Accuracy 83.2% # 447
Number of params 28M # 683
GFLOPs 4.3 # 210
Image Classification ImageNet NAT-Small Top 1 Accuracy 83.7% # 392
Number of params 51M # 790
GFLOPs 7.8 # 278

Methods