Exploring Complementary Strengths of Invariant and Equivariant Representations for Few-Shot Learning

Benchmark Model Rank Results
few-shot-image-classification-on-fc100-5-way-5-shotInvariance-Equivariance#7Accuracy: 65.3
few-shot-image-classification-on-fc100-5-way-1-shotInvariance-Equivariance#9Accuracy: 47.76
few-shot-image-classification-on-cifar-fs-5-way-5-shotInvariance-Equivariance#12Accuracy: 89.74
few-shot-image-classification-on-meta-datasetInvariance-Equivariance#13Accuracy: 68.89
few-shot-image-classification-on-cifar-fs-5-way-1-shotInvariance-Equivariance#15Accuracy: 77.87
few-shot-image-classification-on-tiered-imagenet-5-way-5-shotInvariance-Equivariance#20Accuracy: 87.08
few-shot-image-classification-on-mini-imagenet-5-way-5-shotInvariance-Equivariance#24Accuracy: 84.78
few-shot-image-classification-on-tiered-imagenet-5-way-1-shotInvariance-Equivariance#25Accuracy: 72.21
few-shot-image-classification-on-mini-imagenet-5-way-1-shotInvariance-Equivariance#44Accuracy: 67.28