Inverting visual representations within deep neural networks (DNNs) presents a challenging and important problem in the field of security and privacy for deep learning. The main goal is to invert the features of an unidentified target image generated by a pre-trained DNN, aiming to reconstruct the original image. Feature inversion holds particular significance in understanding the privacy leakage inherent in contemporary split DNN execution techniques, as well as in various applications based on the extracted DNN features. In this paper, we explore the use of diffusion models, a promising technique for image synthesis, to enhance feature inversion quality. We also investigate the potential of incorporating alternative forms of prior knowledge, such as textual prompts and cross-frame temporal correlations, to further improve the quality of inverted features. Our findings reveal that diffusion models can effectively leverage hidden information from the DNN features, resulting in superior reconstruction performance compared to previous methods. This research offers valuable insights into how diffusion models can enhance privacy and security within applications that are reliant on DNN features.
@article{arxiv.2412.10448,
title = {Unlocking Visual Secrets: Inverting Features with Diffusion Priors for Image Reconstruction},
author = {Sai Qian Zhang and Ziyun Li and Chuan Guo and Saeed Mahloujifar and Deeksha Dangwal and Edward Suh and Barbara De Salvo and Chiao Liu},
journal= {arXiv preprint arXiv:2412.10448},
year = {2024}
}