List your company
ResearchFormerly WizWand

Few-shot classification on ChestX (test)

32.1AccuracyP>M>F
19.51622.78326.0529.317Mar 2, 2023Apr 9, 2023May 17, 2023Jun 24, 2023Aug 1, 2023Sep 8, 2023Oct 17, 2023
Updated 8mo ago

Evaluation results

MethodLinks
P>M>FEvaluation protocol=In-Domain [Meta-Training], K-shot=5-shot, N-way=5-way, Backbone=ViT-base2023.1032.1
SimCLRWay=5, Shot=20, Backbone=Conv5, Pre-trained=miniImageNet2023.0329.21
MoCo v2Way=5, Shot=20, Backbone=Conv5, Pre-trained=miniImageNet2023.0329.19
Meta-SVEBMWay=5, Shot=20, Backbone=Conv5, Pre-trained=miniImageNet2023.0328.91
SwAVWay=5, Shot=20, Backbone=Conv5, Pre-trained=miniImageNet2023.0328.57
ProtoDiffway=5-way, shot=5-shot2023.0628.54
PsCoWay=5, Shot=20, Backbone=Conv5, Pre-trained=miniImageNet2023.0327.69
MetaOptEvaluation protocol=Universal Meta-Learning, K-shot=5-shot, N-way=5-way, Backbone frozen=true2023.1027.4
HVMway=5-way, shot=5-shot2023.0627.15
P>M>FEvaluation protocol=In-Domain [Meta-Training], K-shot=1-shot, N-way=5-way, Backbone=ViT-base2023.1027
MetaQDAEvaluation protocol=Universal Meta-Learning, K-shot=5-shot, N-way=5-way2023.1027
Meta-GMVAEWay=5, Shot=20, Backbone=Conv5, Pre-trained=miniImageNet2023.0326.26
Meta-SVEBMWay=5, Shot=5, Backbone=Conv5, Pre-trained=miniImageNet2023.0326.26
ProtoNetEvaluation protocol=Universal Meta-Learning, K-shot=5-shot, N-way=5-way, Backbone frozen=true2023.1025.8
ProtoNetsWay=5, Shot=20, Backbone=Conv5, Pre-trained=miniImageNet2023.0325.72
ProtoNetEvaluation protocol=Universal Meta-Learning, K-shot=5-shot, N-way=5-way, Backbone frozen=false2023.1025.3
MoCo v2Way=5, Shot=5, Backbone=Conv5, Pre-trained=miniImageNet2023.0325.24
SimCLRWay=5, Shot=5, Backbone=Conv5, Pre-trained=miniImageNet2023.0325.14
AFAway=5-way, shot=5-shot2023.0625.02
SwAVWay=5, Shot=5, Backbone=Conv5, Pre-trained=miniImageNet2023.0324.99
PsCoWay=5, Shot=5, Backbone=Conv5, Pre-trained=miniImageNet2023.0324.78
ATAway=5-way, shot=5-shot2023.0624.32
MAMLWay=5, Shot=20, Backbone=Conv5, Pre-trained=miniImageNet2023.0324.25
ProtoNetway=5-way, shot=5-shot2023.0624.05
GNNway=5-way, shot=5-shot2023.0623.87
Meta-GMVAEWay=5, Shot=5, Backbone=Conv5, Pre-trained=miniImageNet2023.0323.23
ProtoNetsWay=5, Shot=5, Backbone=Conv5, Pre-trained=miniImageNet2023.0323.15
ProtoDiffway=5-way, shot=1-shot2023.0623.01
MetaOptEvaluation protocol=Universal Meta-Learning, K-shot=5-shot, N-way=5-way, Backbone frozen=false2023.1023
MetaOptEvaluation protocol=Universal Meta-Learning, K-shot=1-shot, N-way=5-way, Backbone frozen=true2023.1023
HVMway=5-way, shot=1-shot2023.0622.94
AFAway=5-way, shot=1-shot2023.0622.92
ProtoNetEvaluation protocol=Universal Meta-Learning, K-shot=1-shot, N-way=5-way, Backbone frozen=true2023.1022.7
MAMLWay=5, Shot=5, Backbone=Conv5, Pre-trained=miniImageNet2023.0322.61
MetaQDAEvaluation protocol=Universal Meta-Learning, K-shot=1-shot, N-way=5-way2023.1022.6
ProtoNetEvaluation protocol=Universal Meta-Learning, K-shot=1-shot, N-way=5-way, Backbone frozen=false2023.1022.4
ProtoNetway=5-way, shot=1-shot2023.0622.3
CAMLEvaluation protocol=Universal Meta-Learning, K-shot=5-shot, N-way=5-way2023.1022.2
ATAway=5-way, shot=1-shot2023.0622.1
GNNway=5-way, shot=1-shot2023.0621.94
CAMLEvaluation protocol=Universal Meta-Learning, K-shot=1-shot, N-way=5-way2023.1021.5
GPICLEvaluation protocol=Universal Meta-Learning, K-shot=5-shot, N-way=5-way2023.1020.9
MetaOptEvaluation protocol=Universal Meta-Learning, K-shot=1-shot, N-way=5-way, Backbone frozen=false2023.1020.8
SNAILEvaluation protocol=Universal Meta-Learning, K-shot=1-shot, N-way=5-way2023.1020.2
GPICLEvaluation protocol=Universal Meta-Learning, K-shot=1-shot, N-way=5-way2023.1020.1
SNAILEvaluation protocol=Universal Meta-Learning, K-shot=5-shot, N-way=5-way2023.1020