Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks

Benchmark Model Rank Results
category-agnostic-pose-estimation-on-mp100MAML#2Mean PCK@0.2 - 1shot: 61.50
few-shot-image-classification-on-omniglot-5-shot-5-wayMAML#2Accuracy: 99.9
few-shot-image-classification-on-mini-imagenet-cub-5-way-1-shotMAML (Finn et al., 2017)#8Accuracy: 40.15
few-shot-image-classification-on-dirichlet-mini-imagenet-5-way-5-shotMAML#101:1 Accuracy: 64.5
few-shot-image-classification-on-meta-dataset-rankfo-MAML#10Mean Rank: 10.25
few-shot-image-classification-on-mini-imagenet-10-way-5-shotMAML + Transduction#10Accuracy: 48.2
few-shot-image-classification-on-mini-imagenet-10-way-5-shotMAML#13Accuracy: 46.9
few-shot-image-classification-on-omniglot-1-shot-5-wayMAML#10Accuracy: 98.7
few-shot-image-classification-on-tiered-imagenet-10-way-5-shotMAML + Transduction#10Accuracy: 54.7
few-shot-image-classification-on-tiered-imagenet-10-way-5-shotMAML#11Accuracy: 53.3
few-shot-image-classification-on-dirichlet-mini-imagenet-5-way-1-shotMAML#111:1 Accuracy: 47.6
few-shot-image-classification-on-tiered-imagenet-10-way-1-shotMAML + Transduction#11Accuracy: 34.8
few-shot-image-classification-on-tiered-imagenet-10-way-1-shotMAML#12Accuracy: 34.4
few-shot-image-classification-on-mini-imagenet-10-way-1-shotMAML + Transduction#12Accuracy: 31.8
few-shot-image-classification-on-mini-imagenet-10-way-1-shotMAML#13Accuracy: 31.3
few-shot-image-classification-on-meta-datasetfo-MAML#18Accuracy: 57.024
few-shot-image-classification-on-mini-imagenet-5-way-5-shotMAML#88Accuracy: 63.1
few-shot-image-classification-on-mini-imagenet-5-way-1-shotMAML#95Accuracy: 48.7