Screenshots may seem harmless, but hidden invisible watermarks can embed employee IDs or timestamps, survive photo captures, and be extracted to trace the source, as demonstrated with real‑world incidents, demo tools, and even Google’s AI SynthID technology.
Amid rising deep‑fake misuse in entertainment, Ant Group’s AI Security Lab won the CVPR 2026 NTIRE Robust AIGC Image Detection challenge with a ROC AUC of 0.9723, presenting a DINOv3‑based robust detection framework, extensive multi‑source data, and novel augmentation and optimization techniques to combat AI‑generated abuse.
This paper presents a neural‑enhanced camera fingerprint framework combined with zero‑knowledge proof and digital signature schemes to reliably trace the originating device of photos, offering high‑accuracy identification, privacy preservation, and resistance to forgery across various application scenarios.
The ICDAR 2023 Competition, co‑hosted by Alibaba Security and leading Chinese universities, offers a large‑scale e‑commerce text‑tampering dataset of 19,000 images, two challenge tracks (detection and localization), generous prize money up to 100,000 RMB, and a detailed schedule for registration, submissions, and final rankings.
