ImageNet-LT (ImageNet Long-Tailed)

Introduced by Liu et al. in Large-Scale Long-Tailed Recognition in an Open World

ImageNet Long-Tailed is a subset of /dataset/imagenet dataset consisting of 115.8K images from 1000 categories, with maximally 1280 images per class and minimally 5 images per class. The additional classes of images in ImageNet-2010 are used as the open set.

Source: Large-Scale Long-Tailed Recognition in an Open World

Homepage

Benchmarks

Add a new result Link an existing benchmark

Task	Dataset Variant	Best Model
Long-tail Learning	ImageNet-LT	µ2Net+
Long-tail learning with class descriptors	ImageNet-LT-d	DRAGON + Bal'Loss
Conditional Image Generation	ImageNet-LT	StyleGAN2 + NoisyTwins