search

wutong16 / DistributionBalancedLoss

[ ECCV 2020 Spotlight ] Pytorch implementation for "Distribution-Balanced Loss for Multi-Label Classification in Long-Tailed Datasets"
362 stars 46 forks source link

Dataset? #6

Closed PyJulie closed 3 years ago

PyJulie commented 3 years ago

Is the dataset first released by authors? where can I download the dataset you evaluated in your manuscript? Or is there any widely-accepted public dataset for multi-label long-tailed classification task?

wutong16 commented 3 years ago

Yes. The datasets we used are subsets extracted from the original COCO and VOC, and the annotation files are included in the repo under the ./appendix folder. Currently, we're not aware of widely-accepted public datasets for multi-label long-tailed classification, while LVIS is a popular dataset for detection, it's challenging to perform pure classification on it due to its complexity.