Inverse Problems Leveraging Pre-trained Contrastive Representations

Abstract

We study a new family of inverse problems for recovering representations of corrupted data. We assume access to a pre-trained representation learning network R(x) that operates on clean images, like CLIP. The problem is to recover the representation of an image R(x), if we are only given a corrupted version A(x), for some known forward operator A. We propose a supervised inversion method that uses a contrastive objective to obtain excellent representations for highly corrupted images. Using a linear probe on our robust representations, we achieve a higher accuracy than end-to-end supervised baselines when classifying images with various types of distortions, including blurring, additive noise, and random pixel masking. We evaluate on a subset of ImageNet and observe that our method is robust to varying levels of distortion. Our method outperforms end-to-end baselines even with a fraction of the labeled data in a wide range of forward operators.

Citation

APA

Ravula, S., Smyrnis, G., Jordan, M., & Dimakis, A. G. (2021). Inverse Problems Leveraging Pre-trained Contrastive Representations. Advances in Neural Information Processing Systems, 34, 8753-8765.

Bibtex

@article{ravula2021inverse,

title={Inverse Problems Leveraging Pre-trained Contrastive Representations},

author={Ravula, Sriram and Smyrnis, Georgios and Jordan, Matt and Dimakis, Alexandros G},

journal={Advances in Neural Information Processing Systems},

volume={34},

pages={8753--8765},

year={2021}

}