List your company
ResearchFormerly WizWand

Cross-domain Few-shot Character Recognition on Omniglot to EMNIST (novel classes)

88.94AccuracyRelationNet
62.9469.6976.4483.19Apr 8, 2019
Updated 8mo ago

Evaluation results

MethodLinks
RelationNetN-shot=5-shot, Backbone=Conv-4, Data Augmentation=None, N-way=5-way2019.0488.94
MAMLN-shot=5-shot, Backbone=Conv-4, Data Augmentation=None, N-way=5-way2019.0488.24
MatchingNetN-shot=5-shot, Backbone=Conv-4, Data Augmentation=None, N-way=5-way2019.0487.6
Baseline++N-shot=5-shot, Backbone=Conv-4, Data Augmentation=None, N-way=5-way, Training Epochs=52019.0487.31
ProtoNetN-shot=5-shot, Backbone=Conv-4, Data Augmentation=None, N-way=5-way2019.0487.04
BaselineN-shot=5-shot, Backbone=Conv-4, Data Augmentation=None, N-way=5-way, Training Epochs=52019.0486
RelationNetN-shot=1-shot, Backbone=Conv-4, Data Augmentation=None, N-way=5-way2019.0475.55
MatchingNetN-shot=1-shot, Backbone=Conv-4, Data Augmentation=None, N-way=5-way2019.0472.71
MAMLN-shot=1-shot, Backbone=Conv-4, Data Augmentation=None, N-way=5-way2019.0472.04
ProtoNetN-shot=1-shot, Backbone=Conv-4, Data Augmentation=None, N-way=5-way2019.0470.43
Baseline++N-shot=1-shot, Backbone=Conv-4, Data Augmentation=None, N-way=5-way, Training Epochs=52019.0464.74
BaselineN-shot=1-shot, Backbone=Conv-4, Data Augmentation=None, N-way=5-way, Training Epochs=52019.0463.94