Key points are not available for this paper at this time.
We propose a novel method for privacy-preserving fine-tuning vision transformers (ViTs) with encrypted images. Conventional methods using encrypted images degrade model performance compared with that of using plain images due to the influence of image encryption. In contrast, the proposed encryption method using restricted random permutation matrices can provide a higher performance than the conventional ones.
Building similarity graph...
Analyzing shared references across papers
Loading...
Horio et al. (Fri,) studied this question.
synapsesocial.com/papers/68e5c0e5b6db64358755854d — DOI: https://doi.org/10.48550/arxiv.2408.08529
K. Horio
Kiyoshi Nishikawa
Tokyo Metropolitan University
Hitoshi Kiya
Tokyo Metropolitan University
Building similarity graph...
Analyzing shared references across papers
Loading...