Escaping Saddle Points for Effective Generalization on Class-Imbalanced Data

Open paper
Benchmark Model Rank Results
long-tail-learning-on-cifar-10-lt-r-10LDAM + DRW + SAM#23Error Rate: 10.6
long-tail-learning-on-cifar-10-lt-r-100GLMC + SAM#2Error Rate: 10.82
long-tail-learning-on-cifar-10-lt-r-100VS + SAM#16Error Rate: 17.6
long-tail-learning-on-cifar-10-lt-r-50GLMC + SAM#1Error Rate: 8.44
long-tail-learning-on-cifar-100-lt-r-100GLMC + SAM#6Error Rate: 40.99
long-tail-learning-on-cifar-100-lt-r-100PaCo + SAM#20Error Rate: 47.0
long-tail-learning-on-cifar-100-lt-r-100VS + SAM#37Error Rate: 53.4
long-tail-learning-on-cifar-100-lt-r-50GLMC + SAM#5Error Rate: 34.72
long-tail-learning-on-imagenet-ltLDAM + DRW + SAM#43Top-1 Accuracy: 53.1
long-tail-learning-on-inaturalist-2018LDAM + DRW + SAM#32Top-1 Accuracy: 70.1

Rank counts only results with a code link.