Fooling Downstream Classifiers via Attacking Contrastive Learning Pre-trained Models.

Chenggang Li, Anjie Peng, Hui Zeng, Kaijun Wu, Wenxin Yu

Browse the full ICONIP paper archive.