BatchFormer: Learning to Explore Sample Relationships for Robust Representation Learning

Open paper
Benchmark Model Rank Results
domain-generalization-on-pacs-2BatchFormer(ResNet-50, SWAD)#23Average Accuracy: 88.6
long-tail-learning-on-cifar-100-lt-r-100Paco + BatchFormer#22Error Rate: 47.6
long-tail-learning-on-cifar-100-lt-r-100Balanced + BatchFormer#23Error Rate: 48.3
long-tail-learning-on-imagenet-ltBatchFormer(ResNet-50, PaCo)#27Top-1 Accuracy: 57.4
long-tail-learning-on-imagenet-ltBatchFormer(ResNet-50, RIDE)#34Top-1 Accuracy: 55.7
long-tail-learning-on-inaturalist-2018BatchFormer(ResNet-50, RIDE)#19Top-1 Accuracy: 74.1%

Rank counts only results with a code link.