Be Careful What You Smooth For: Label Smoothing Can Be a Privacy Shield but Also a Catalyst for Model Inversion Attacks

Lukas Struppek, Dominik Hintersdorf, Kristian Kersting

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.

Keywords: model inversion attacks, privacy, label smoothing, defense

Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2024/AuthorGuide.

TL;DR: Label smoothing regularization can reinforce a model's privacy leakage but can also act as a practical and powerful defense mechanism.

Abstract: Label smoothing – using softened labels instead of hard ones – is a widely adopted regularization method for deep learning, showing diverse benefits such as enhanced generalization and calibration. Its implications for preserving model privacy, however, have remained unexplored. To fill this gap, we investigate the impact of label smoothing on model inversion attacks (MIAs), which aim to generate class-representative samples by exploiting the knowledge encoded in a classifier, thereby inferring sensitive information about its training data. Through extensive analyses, we uncover that traditional label smoothing fosters MIAs, thereby increasing a model's privacy leakage. Even more, we reveal that smoothing with negative factors counters this trend, impeding the extraction of class-related information and leading to privacy preservation, beating state-of-the-art defenses. This establishes a practical and powerful novel way for enhancing model resilience against MIAs.

Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors' identity.

No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.

Primary Area: societal considerations including fairness, safety, privacy

Submission Number: 1183