Rethinking Visual Provenance: Detection and Watermarking Across Direct Visual Generation and LLM-Driven Code Rendering
We develop a production-centered framework that compares detection and watermarking across both routes.
Key points
- AI systems create images and videos with image/video generation models or by writing code and graphics descriptions that are then rendered.
- An explicit verification specification distinguishes passive inference, message recovery, and authenticated provenance.
- We organize image, video, source-code, and rendering-aware watermarks by production stage.
- We examine the different requirements of generated images and video, plots and SVG, programmable video, and agent-composed workflows.
Sources (1)
- [1]Rethinking Visual Provenance: Detection and Watermarking Across Direct Visual Generation and LLM-Driven Code RenderingarXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 6, 10:53 AM
We develop a production-centered framework that compares detection and watermarking across both routes.
AI systems create images and videos with image/video generation models or by writing code and graphics descriptions that are then rendered.
Extractive summary: sentences quoted from the sources.