Next-ViT: Next Generation Vision Transformer for Efficient Deployment in Realistic Industrial Scenarios

Open paper
Benchmark Model Rank Results
image-classification-on-imagenetNext-ViT-L @384#292Top 1 Accuracy: 84.7%Number of params: 57.8MGFLOPs: 32
image-classification-on-imagenetNext-ViT-B#445Top 1 Accuracy: 83.2%Number of params: 44.8MGFLOPs: 8.3
image-classification-on-imagenetNext-ViT-S#516Top 1 Accuracy: 82.5%Number of params: 31.7MGFLOPs: 5.8

Rank counts only results with a code link.