Investigating and Mitigating the Side Effects of Noisy Views for Self-Supervised Clustering Algorithms in Practical Multi-View Scenarios

The public datasets and our trained models are available at Download or 国内下载源.

Run the code by: python run.py

Train the new model: Test = False, in run.py

Test the trained model: Test = True, in run.py

Requirement:

python==3.7.11

numpy==1.20.1

scikit-learn==0.22.2.post1

scipy==1.6.2

pytorch==1.9.0