CleanCLIP: Mitigating Data Poisoning Attacks in Multimodal Contrastive Learning

Hritik Bansal, Nishad Singhi, Yu Yang, Fan Yin, Aditya Grover, and Kai-Wei Chang, in ICLR workshop on Reliable and Trustworthy Large Scale Machine Learning, 2023.

Abstract

Multimodal contrastive pretraining has been used to train multimodal representation models, such as CLIP, on large amounts of paired image-text data. However, previous studies have revealed that such models are vulnerable to backdoor attacks. Specifically, when trained on backdoored examples, CLIP learns spurious correlations between the embedded backdoor trigger and the target label, aligning their representations in the joint embedding space. Injecting even a small number of poisoned examples, such as 75 examples in 3 million pretraining data, can significantly manipulate the model’s behavior, making it difficult to detect or unlearn such correlations. To address this issue, we propose CleanCLIP, a finetuning framework that weakens the learned spurious associations introduced by backdoor attacks by independently re-aligning the representations for individual modalities. We demonstrate that unsupervised finetuning using a combination of multimodal contrastive and unimodal self-supervised objectives for individual modalities can significantly reduce the impact of the backdoor attack. We show empirically that CleanCLIP maintains model performance on benign examples while erasing a range of backdoor attacks on multimodal contrastive learning.

Bib Entry

@inproceedings{bansal2023cleanclip,
  author = {Bansal, Hritik and Singhi, Nishad and Yang, Yu and Yin, Fan and Grover, Aditya and Chang, Kai-Wei},
  title = {CleanCLIP: Mitigating Data Poisoning Attacks in Multimodal Contrastive Learning},
  booktitle = {ICLR workshop on Reliable and Trustworthy Large Scale Machine Learning},
  year = {2023}
}

CleanCLIP: Mitigating Data Poisoning Attacks in Multimodal Contrastive Learning