PyTorch implementation for Robust Multi-view Clustering with Incomplete Information (TPAMI 2022).

pytorch>=1.5.0

numpy>=1.18.2

scikit-learn>=0.22.2

munkres>=1.1.2

logging>=0.5.1.2

The Scene-15 and Reuters-dim10 datasets are placed in "datasets" folder. Another datasets could be downloaded from Google cloud or Baidu cloud with password "rqv4".

Train a model with different settings

# Partially Aligned, Table 1, Table 6
python run.py --data 0 --gpu 0 --settings 0 --aligned-prop 0.5 --complete-prop 1.0
# Fully Aligned, Table 1, Table 6
python run.py --data 0 --gpu 0 --settings 0 --aligned-prop 1.0 --complete-prop 1.0
# Incomplete, Table 5, Table 6
python run.py --data 0 --gpu 0 --settings 1 --aligned-prop 1.0 --complete-prop 0.5
# Complete, Table 5, Table 6
python run.py --data 0 --gpu 0 --settings 1 --aligned-prop 1.0 --complete-prop 1.0
# PVP + PSP, Figure 9, Table 6
python run.py --data 0 --gpu 0 --settings 2 --aligned-prop 0.5 --complete-prop 0.5

• --data: choice of datasets.

• --gpu: which gpu to run.

• --settings: 0-PVP, 1-PSP, 2-Both.

• --aligned-prop: known aligned proportions for training.

• --complete-prop: known complete proportions for training.

  Parameters: More parameters and descriptions can be found in the script.

  Training Log: The training log will be saved in log/

If you find our work useful in your research, please consider citing:

@inproceedings{yang2021MvCLN,
   title={Partially View-aligned Representation Learning with Noise-robust Contrastive Loss},
   author={Mouxing Yang, Yunfan Li, Zhenyu Huang, Zitao Liu, Peng Hu, Xi Peng},
   booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
   month={June},
   year={2021}
}
@article{yang2022SURE,
	title={Robust Multi-view Clustering with Incomplete Information},
  	author={Yang, Mouxing and Li, Yunfan and Hu, Peng and Bai, Jinfeng and Lv, Jian Cheng and Peng, Xi},  
	journal={IEEE Transactions on Pattern Analysis and Machine Intelligence},     
 	year={2022},  
}