VL-LTR: Learning Class-wise Visual-Linguistic Representation for Long-Tailed Visual Recognition

Open paper
Benchmark Model Rank Results
long-tail-learning-on-inaturalist-2018VL-LTR (ViT-B-16)#4Top-1 Accuracy: 81.0%
long-tail-learning-on-inaturalist-2018VL-LTR (ResNet-50)#16Top-1 Accuracy: 74.6%
long-tail-learning-on-places-ltVL-LTR (ViT-B-16)#4Top-1 Accuracy: 50.1
long-tail-learning-on-places-ltVL-LTR (ResNet-50)#7Top-1 Accuracy: 48.0
long-tail-learning-on-imagenet-ltVL-LTR (ViT-B-16)#5Top-1 Accuracy: 77.2
long-tail-learning-on-imagenet-ltVL-LTR (ResNet-50)#9Top-1 Accuracy: 70.1
image-classification-on-inaturalist-2018VL-LTR (ViT-B-16)#11Top-1 Accuracy: 81.0%
image-classification-on-inaturalist-2018VL-LTR (ResNet-50)#26Top-1 Accuracy: 74.6%

Rank counts only results with a code link.