BLT: Balancing Long-Tailed Datasets with Adversarially-Perturbed Images
Open paper
Benchmark
Model
Rank
Results
long-tail-learning-on-imagenet-lt
BLT (DenseNet-121)
#56
Top-1 Accuracy: 44.7
long-tail-learning-on-imagenet-lt
BLT (ResNet-10)
#64
Top-1 Accuracy: 38.0
Rank counts only results with a code link.