WAVES: Benchmarking the Robustness of Image Watermarks

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: An, Bang, Ding, Mucong, Rabbani, Tahseen, Agrawal, Aakriti, Xu, Yuancheng, Deng, Chenghao, Zhu, Sicheng, Mohamed, Abdirisak, Wen, Yuxin, Goldstein, Tom, Huang, Furong
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866914828767985664
author An, Bang
Ding, Mucong
Rabbani, Tahseen
Agrawal, Aakriti
Xu, Yuancheng
Deng, Chenghao
Zhu, Sicheng
Mohamed, Abdirisak
Wen, Yuxin
Goldstein, Tom
Huang, Furong
author_facet An, Bang
Ding, Mucong
Rabbani, Tahseen
Agrawal, Aakriti
Xu, Yuancheng
Deng, Chenghao
Zhu, Sicheng
Mohamed, Abdirisak
Wen, Yuxin
Goldstein, Tom
Huang, Furong
contents In the burgeoning age of generative AI, watermarks act as identifiers of provenance and artificial content. We present WAVES (Watermark Analysis Via Enhanced Stress-testing), a benchmark for assessing image watermark robustness, overcoming the limitations of current evaluation methods. WAVES integrates detection and identification tasks and establishes a standardized evaluation protocol comprised of a diverse range of stress tests. The attacks in WAVES range from traditional image distortions to advanced, novel variations of diffusive, and adversarial attacks. Our evaluation examines two pivotal dimensions: the degree of image quality degradation and the efficacy of watermark detection after attacks. Our novel, comprehensive evaluation reveals previously undetected vulnerabilities of several modern watermarking algorithms. We envision WAVES as a toolkit for the future development of robust watermarks. The project is available at https://wavesbench.github.io/
format Preprint
id arxiv_https___arxiv_org_abs_2401_08573
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle WAVES: Benchmarking the Robustness of Image Watermarks
An, Bang
Ding, Mucong
Rabbani, Tahseen
Agrawal, Aakriti
Xu, Yuancheng
Deng, Chenghao
Zhu, Sicheng
Mohamed, Abdirisak
Wen, Yuxin
Goldstein, Tom
Huang, Furong
Computer Vision and Pattern Recognition
Cryptography and Security
Machine Learning
In the burgeoning age of generative AI, watermarks act as identifiers of provenance and artificial content. We present WAVES (Watermark Analysis Via Enhanced Stress-testing), a benchmark for assessing image watermark robustness, overcoming the limitations of current evaluation methods. WAVES integrates detection and identification tasks and establishes a standardized evaluation protocol comprised of a diverse range of stress tests. The attacks in WAVES range from traditional image distortions to advanced, novel variations of diffusive, and adversarial attacks. Our evaluation examines two pivotal dimensions: the degree of image quality degradation and the efficacy of watermark detection after attacks. Our novel, comprehensive evaluation reveals previously undetected vulnerabilities of several modern watermarking algorithms. We envision WAVES as a toolkit for the future development of robust watermarks. The project is available at https://wavesbench.github.io/
title WAVES: Benchmarking the Robustness of Image Watermarks
topic Computer Vision and Pattern Recognition
Cryptography and Security
Machine Learning
url https://arxiv.org/abs/2401.08573