Squeezing Backbone Feature Distributions to the Max for Efficient Few-Shot Learning

Benchmark Model Rank Results
few-shot-image-classification-on-cifar-fs-5PEMnE-BMS*#3Accuracy: 88.44
few-shot-image-classification-on-cifar-fs-5-1PEMnE-BMS*#5Accuracy: 91.86
few-shot-image-classification-on-cub-200-5PEMnE-BMS*#4Accuracy: 96.43
few-shot-image-classification-on-cub-200-5-1PEMnE-BMS*#5Accuracy: 94.78
few-shot-image-classification-on-mini-2PEMnE-BMS* (transductive)#7Accuracy: 85.54
few-shot-image-classification-on-mini-2PEMbE-NCM (inductive)#36Accuracy: 68.43
few-shot-image-classification-on-mini-3PEMnE-BMS*(transductive)#5Accuracy: 91.53
few-shot-image-classification-on-mini-3PEMbE-NCM (inductive)#25Accuracy: 84.67
few-shot-image-classification-on-mini-5PEMnE-BMS*#2Accuracy: 63.90
few-shot-image-classification-on-mini-6PEMnE-BMS#2Accuracy: 79.15
few-shot-image-classification-on-tieredPEMnE-BMS*#3Accuracy: 86.07
few-shot-image-classification-on-tiered-1PEMnE-BMS*#4Accuracy: 91.09