arXiv：cs.AI· Saibo Ye, Huajie Chen, Xin Guo, Le Yang, Chi Liu, Xiangyu Hu, Jingjing Guo, Tianqing Zhu·· 4 小时前AI 评分33
LiBRA：基于双向潜空间优化的检测感知图像水印去除
LiBRA: Detection-Aware Image Watermark Removal via Bidirectional Latent Optimization
AI 导读
LiBRA（Latent In-band Bidirectional Removal Attack）通过公开自编码器的潜空间有界修改，将水印解码置信度从两个方向引导至随机猜测水平，使水印不可检测同时保持图像质量。该方法在已知水印密钥与解码器条件下，避免产生仍可被检测的倒置水印，并用精确双侧二项检验验证去除效果。
正文
Abstract:Digital watermarking supports source attribution for AI-generated images, but its reliability depends on resistance to removal attacks. Some attacks attempt to remove watermarks by forcing the decoded watermark to differ from the original. However, this can produce an inverted watermark that remains detectable, causing removal to fail, while further attempts to alter the watermark may unnecessarily degrade image quality. To address these limitations, we present LiBRA (Latent In-band Bidirectional Removal Attack), which aims to make watermarks undetectable while preserving image quality. Instead of continually pushing the watermark toward inversion, LiBRA adjusts the image to conceal the watermark without encouraging further changes that could degrade image quality. Some attacks keep pushing decoded bits away from the original watermark, even when further changes preserve detectability and damage image quality. With access to the watermark key and decoder, LiBRA makes bounded changes in a public autoencoder's latent space. Unlike inversion-driven objectives that cannot correct excessive inversion, LiBRA guides average decoding confidence toward random guessing from either direction. This helps avoid an inverted but detectable watermark. Leaving individual bits flexible allows image-quality constraints to favor less damaging changes, while an optional frequency-guided mask limits their location. We verify removal using an exact two-sided binomial test rather than assuming the confidence target guarantees success.
| Subjects: | Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.03166 [cs.CR] |
| (or arXiv:2610.03166v1 [cs.CR] for this version) | |
| https://doi.org/10.48550/arXiv.2610.03166 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Huajie Chen [view email]
[v1]
Fri, 2 Oct 2026 11:44:09 UTC (6,383 KB)
来源：arXiv：cs.AI · arxiv.org