Prototypical Networks for Few-shot Learning

Benchmark Model Rank Results
category-agnostic-pose-estimation-on-mp100ProtoNet#3Mean PCK@0.2 - 1shot: 44.78
few-shot-image-classification-on-dirichletProtoNet#91:1 Accuracy: 53.6
few-shot-image-classification-on-dirichlet-1ProtoNet#71:1 Accuracy: 74.2
few-shot-image-classification-on-meta-datasetPrototypical Networks#16Accuracy: 60.573
few-shot-image-classification-on-meta-dataset-1Prototypical Networks#8Mean Rank: 8.5
few-shot-image-classification-on-mini-12Prototypical Networks (Higher Way)#9Accuracy: 34.6
few-shot-image-classification-on-mini-12Prototypical Networks#10Accuracy: 32.9
few-shot-image-classification-on-mini-13Prototypical Networks (Higher Way)#8Accuracy: 50.1
few-shot-image-classification-on-mini-13Prototypical Networks#9Accuracy: 49.3
few-shot-image-classification-on-mini-2Prototypical Networks#92Accuracy: 49.42
few-shot-image-classification-on-mini-3Prototypical Networks#79Accuracy: 68.20
few-shot-image-classification-on-mini-5ProtoNet (Snell et al., 2017)#5Accuracy: 45.31
few-shot-image-classification-on-omniglot-1-1Prototypical Networks#11Accuracy: 96%
few-shot-image-classification-on-omniglot-1-2Prototypical Networks#9Accuracy: 98.8
few-shot-image-classification-on-omniglot-5-1Prototypical Networks#10Accuracy: 98.9%
few-shot-image-classification-on-omniglot-5-2Prototypical Networks#9Accuracy: 99.7
few-shot-image-classification-on-stanford-1Prototypical Nets++#3Accuracy: 48.19
few-shot-image-classification-on-stanford-2Prototypical Nets++#3Accuracy: 40.90
few-shot-image-classification-on-stanford-3Prototypical Nets++#3Accuracy: 52.93
few-shot-image-classification-on-tiered-2Prototypical Networks (Higher Way)#7Accuracy: 38.6
few-shot-image-classification-on-tiered-2Prototypical Networks#8Accuracy: 37.3
few-shot-image-classification-on-tiered-3Prototypical Networks (Higher Way)#6Accuracy: 58.3
few-shot-image-classification-on-tiered-3Prototypical Networks#9Accuracy: 57.8