Visual Prompt Tuning

Benchmark Model Rank Results
long-tail-learning-on-cifar-100-lt-r-10VPT#3Error Rate: 10.4
long-tail-learning-on-cifar-100-lt-r-100VPT#4Error Rate: 19
long-tail-learning-on-cifar-100-lt-r-50VPT#3Error Rate: 15.2
visual-prompt-tuning-on-fgvcVPT-Deep(ViT-B/16_MoCo_v3_pretrained_ImageNet-1K)#1Mean Accuracy: 83.12
visual-prompt-tuning-on-fgvcVPT-Shallow (ViT-B/16_MoCo_v3_pretrained_ImageNet-1K)#3Mean Accuracy: 79.26
visual-prompt-tuning-on-fgvcVPT-Deep (ViT-B/16_MAE_pretrained_ImageNet-1K)#5Mean Accuracy: 72.02
visual-prompt-tuning-on-fgvcVPT-Shallow (ViT-B/16_MAE_pretrained_ImageNet-1K)#6Mean Accuracy: 57.84
visual-prompt-tuning-on-vtab-1k-natural-7VPT-Deep(ViT-B/16_MoCo_v3_pretrained_ImageNet-1K)#2Mean Accuracy: 70.27
visual-prompt-tuning-on-vtab-1k-natural-7VPT-Shallow(ViT-B/16_MoCo_v3_pretrained_ImageNet-1K)#3Mean Accuracy: 67.34
visual-prompt-tuning-on-vtab-1k-natural-7VPT-Shallow(ViT-B/16_MAE_pretrained_ImageNet-1K)#5Mean Accuracy: 39.96
visual-prompt-tuning-on-vtab-1k-natural-7VPT-Deep(ViT-B/16_MAE_pretrained_ImageNet-1K)#6Mean Accuracy: 36.02
visual-prompt-tuning-on-vtab-1k-specialized-4VPT-Deep(ViT-B/16_MoCo_v3_pretrained_ImageNet-1K)#2Mean Accuracy: 83.04
visual-prompt-tuning-on-vtab-1k-specialized-4VPT-Shallow(ViT-B/16_MoCo_v3_pretrained_ImageNet-1K)#3Mean Accuracy: 82.26
visual-prompt-tuning-on-vtab-1k-specialized-4VPT-Shallow(ViT-B/16_MAE_pretrained_ImageNet-1K)#5Mean Accuracy: 69.65
visual-prompt-tuning-on-vtab-1k-specialized-4VPT-Deep(ViT-B/16_MAE_pretrained_ImageNet-1K)#6Mean Accuracy: 60.61
visual-prompt-tuning-on-vtab-1k-structured-8VPT-Deep(ViT-B/16_MoCo_v3_pretrained_ImageNet-1K)#2Mean Accuracy: 42.38
visual-prompt-tuning-on-vtab-1k-structured-8VPT-Shallow(ViT-B/16_MoCo_v3_pretrained_ImageNet-1K)#3Mean Accuracy: 37.55
visual-prompt-tuning-on-vtab-1k-structured-8VPT-Shallow(ViT-B/16_MAE_pretrained_ImageNet-1K)#5Mean Accuracy: 27.50
visual-prompt-tuning-on-vtab-1k-structured-8VPT-Deep(ViT-B/16_MAE_pretrained_ImageNet-1K)#6Mean Accuracy: 26.57