{"ID":2859030,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2510.07538","arxiv_id":"2510.07538","title":"Low-Compute Watermark Removal via Dual-Domain Natural Projection","abstract":"Effective removal of semantic watermarks requires balancing three competing objectives: \\emph{high removal success}, \\emph{low perceptual distortion}, and \\emph{low computational cost}. However, existing single-image attacks typically optimize only for the first two, achieving strong watermark suppression but relying on expensive, multi-step optimization that limits practical deployment. In this work, we show that this trade-off is fundamental: no current approach achieves all three properties simultaneously. We introduce \\textsc{DAWN}, a lightweight, training-free attack that explicitly targets the low-cost regime while maintaining competitive removal performance. \\textsc{DAWN} works by projecting a watermarked image onto natural-image priors in complementary frequency and semantic spaces, suppressing watermark signals that deviate from natural statistics, and then applying a decoupled perceptual-alignment step to restore visual consistency with minimal artifact. Across diverse pixel-, frequency-, and latent-space watermarking schemes, \\textsc{DAWN} consistently reduces detectability while preserving structural and semantic fidelity, demonstrating that efficient, low-resource watermark removal is feasible with only modest perceptual degradation. Our code is available at https://github.com/Pragati-Meshram/DAWN.","short_abstract":"Effective removal of semantic watermarks requires balancing three competing objectives: \\emph{high removal success}, \\emph{low perceptual distortion}, and \\emph{low computational cost}. However, existing single-image attacks typically optimize only for the first two, achieving strong watermark suppression but relying o...","url_abs":"https://arxiv.org/abs/2510.07538","url_pdf":"https://arxiv.org/pdf/2510.07538v2","authors":"[\"Pragati Shuddhodhan Meshram\",\"Varun Chandrasekaran\"]","published":"2025-10-08T20:54:22Z","proceeding":"cs.CV","tasks":"[\"cs.CV\"]","methods":"[]","has_code":false,"code_links":[{"ID":608606,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_id":2859030,"paper_url":"https://arxiv.org/abs/2510.07538","paper_title":"Low-Compute Watermark Removal via Dual-Domain Natural Projection","repo_url":"https://github.com/Pragati-Meshram/DAWN","is_official":false,"mentioned_in_paper":false,"mentioned_in_github":true,"github_stars":0}]}