Instance Credibility Inference for Few-Shot Learning

Benchmark Model Rank Results
few-shot-image-classification-on-dirichlet-tiered-imagenet-5-way-1-shotLR+ICI#11:1 Accuracy: 74.6
few-shot-image-classification-on-dirichlet-tiered-imagenet-5-way-5-shotLR+ICI#31:1 Accuracy: 85.1
few-shot-image-classification-on-tiered-imagenet-5-way-1-shotICI#6Accuracy: 84.01
few-shot-image-classification-on-dirichlet-mini-imagenet-5-way-1-shotLR-ICI#71:1 Accuracy: 58.7
few-shot-image-classification-on-dirichlet-mini-imagenet-5-way-5-shotLR-ICI#81:1 Accuracy: 73.5
few-shot-image-classification-on-cub-200-5-way-1-shotICI#11Accuracy: 89.58
few-shot-image-classification-on-cub-200-5-way-5-shotICI#11Accuracy: 92.48
few-shot-image-classification-on-tiered-imagenet-5-way-5-shotICI#11Accuracy: 89.00
few-shot-image-classification-on-cifar-fs-5-way-1-shotICI#20Accuracy: 76.51
few-shot-image-classification-on-cifar-fs-5-way-5-shotICI#30Accuracy: 84.32
few-shot-image-classification-on-mini-imagenet-5-way-1-shotICI#35Accuracy: 69.66
few-shot-image-classification-on-mini-imagenet-5-way-5-shotICI#50Accuracy: 80.11