CLIP is a widely used foundational vision-language model that is used for zero-shot image recognition and other image-text alignment tasks. We demonstrate that CLIP is vulnerable to change in image quality under compression. This surprising result is further analysed using an attribution method-Integrated Gradients. Using this attribution method, we are able to better understand both quantitatively and qualitatively exactly the nature in which the compression affects the zero-shot recognition accuracy of this model. We evaluate this extensively on CIFAR-10 and STL-10. Our work provides the basis to understand this vulnerability of CLIP and can help us develop more effective methods to improve the robustness of CLIP and other vision-language models.
@article{arxiv.2311.14029,
title = {Understanding the Vulnerability of CLIP to Image Compression},
author = {Cangxiong Chen and Vinay P. Namboodiri and Julian Padget},
journal= {arXiv preprint arXiv:2311.14029},
year = {2023}
}
Comments
R0-FoMo: Workshop on Robustness of Few-shot and Zero-shot Learning in Foundation Models at NeurIPS 2023