Advancing Vision Transformer with Enhanced Spatial Priors

Open paper
Benchmark Model Rank Results
image-classification-on-imagenetEVT-B–Top 1 Accuracy: 85.3%GFLOPs: 9.8
image-classification-on-imagenetEVT-L–Top 1 Accuracy: 85.8%GFLOPs: 18.2
image-classification-on-imagenetEVT-S–Top 1 Accuracy: 84.4%GFLOPs: 4.6
image-classification-on-imagenetEVT-T–Top 1 Accuracy: 83.0%GFLOPs: 2.5
image-classification-on-imagenetEVT-XL–Top 1 Accuracy: 86.3%GFLOPs: 36.4
object-detection-on-coco-val2017EVT-B–box AP: 55.5AP50: 74.0AP75: 60.2
object-detection-on-coco-val2017EVT-L–box AP: 55.8AP50: 74.3AP75: 60.4
object-detection-on-coco-val2017EVT-S–box AP: 53.8AP50: 72.4AP75: 58.4

Rank counts only results with a code link.