Dist-PU: Positive and Unlabeled Learning From a Label Disrtibution Perspective

This is a Pytorch implementation of Dist-PU.

Environment

GPU:

• Geoforce RTX 3090
• cuda 11.1

OS:

• ubuntu 18.04.5

Python Related:

• python 3.7
• pytorch 1.8.1
• torchvision 0.9.1
• numpy 1.19.2
• sklearn 0.24.1

Data Preparation

1. Download CIFAR-10 python version (163MB) from http://www.cs.utoronto.ca/~kriz/cifar.html to your machine.
2. Decompress the downloaded file cifar-10-python.tar.gz from the first step.
3. Usually the second step would result in a new directory like '*/cifar-10-batches-py/' with files in it including:

• data_batch_[1-5]
• test_batch
• batches.meta
• readme.html

Command

python train.py --device GPUID --datapath DATAPATH

bibtex

@InProceedings{Zhao_2022_CVPR, author = {Zhao, Yunrui and Xu, Qianqian and Jiang, Yangbangyan and Wen, Peisong and Huang, Qingming}, title = {Dist-PU: Positive-Unlabeled Learning From a Label Distribution Perspective}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}, month = {June}, year = {2022}, pages = {14461-14470} }