LocalViT: Bringing Locality to Vision Transformers

Open paper
Benchmark Model Rank Results
image-classification-on-imagenetLocalViT-S#721Top 1 Accuracy: 80.8%Number of params: 22.4MGFLOPs: 4.6
image-classification-on-imagenetLocalViT-PVT#883Top 1 Accuracy: 78.2%Number of params: 13.5MGFLOPs: 4.8
image-classification-on-imagenetLocalViT-TNT#969Top 1 Accuracy: 75.9%Number of params: 6.3MGFLOPs: 1.4
image-classification-on-imagenetLocalViT-T#1010Top 1 Accuracy: 74.8%Number of params: 5.9MGFLOPs: 1.3
image-classification-on-imagenetLocalViT-T2T#1047Top 1 Accuracy: 72.5%Number of params: 4.3MGFLOPs: 1.2

Rank counts only results with a code link.