Going deeper with Image Transformers

Open paper
Benchmark Model Rank Results
image-classification-on-stanford-carsCaiT-M-36 U 224#4Accuracy: 94.2
image-classification-on-cifar-10CaiT-M-36 U 224#6Percentage correct: 99.4
image-classification-on-inaturalist-2019CaiT-M-36 U 224#9Top-1 Accuracy: 81.8
image-classification-on-cifar-100CaiT-M-36 U 224#14Percentage correct: 93.1
image-classification-on-flowers-102CaiT-M-36 U 224#19Accuracy: 99.1
image-classification-on-inaturalist-2018CaiT-M-36 U 224#19Top-1 Accuracy: 78%
image-classification-on-imagenet-realCAIT-M36-448#23Accuracy: 90.2%
image-classification-on-imagenet-v2CAIT-M36-448#27Top 1 Accuracy: 76.7
image-classification-on-imagenetCaiT-M-48-448#146Top 1 Accuracy: 86.5%Number of params: 438MGFLOPs: 377.3
image-classification-on-imagenetCAIT-M36-448#167Top 1 Accuracy: 86.3%Number of params: 271MGFLOPs: 247.8
image-classification-on-imagenetCAIT-M-36#185Top 1 Accuracy: 86.1%Number of params: 270.9MGFLOPs: 173.3
image-classification-on-imagenetCAIT-M-24#203Top 1 Accuracy: 85.8%Number of params: 185.9MGFLOPs: 116.1
image-classification-on-imagenetCAIT-S-36#244Top 1 Accuracy: 85.4%Number of params: 68.2MGFLOPs: 48
image-classification-on-imagenetCAIT-S-48#254Top 1 Accuracy: 85.3%Number of params: 89.5MGFLOPs: 63.8
image-classification-on-imagenetCAIT-S-24#273Top 1 Accuracy: 85.1%Number of params: 46.9MGFLOPs: 32.2
image-classification-on-imagenetCAIT-XS-36#303Top 1 Accuracy: 84.8%Number of params: 38.6MGFLOPs: 28.8
image-classification-on-imagenetCAIT-XS-24#375Top 1 Accuracy: 84.1%Number of params: 26.6MGFLOPs: 19.3
image-classification-on-imagenetCAIT-XXS-36#601Top 1 Accuracy: 82.2%Number of params: 17.3MGFLOPs: 14.3
image-classification-on-imagenetCAIT-XXS-24#715Top 1 Accuracy: 80.9%Number of params: 12MGFLOPs: 9.6

Rank counts only results with a code link.