FLock.io (@flock_io): FLock's new research paper — "Identity-Decoupled Anonymization for Visual Evidence in Multi-modal Retrieval-Augmented Generation" — has been accepted…
Estimated coin impact
Scores are not price-change percentages. High impact means a score of +70 or more, or -70 or less.
Impact unavailable
FLock's new research paper — "Identity-Decoupled Anonymization for Visual Evidence in Multi-modal Retrieval-Augmented Generation" — has been accepted to ACM ICMR 2026! 🎉(@TheOfficialACM) Multi-modal RAG systems retrieve visual evidence from large image corpora to ground AI responses — but those images frequently contain human faces, which are considered sensitive personal data under GDPR and CCPA. Existing anonymization methods either destroy the visual detail the AI needs for reasoning or fail to provide strong privacy guarantees. Our paper introduces a framework that disentangles identity from other features in latent space, allowing AI to analyze images without recognizing who’s in them. The result is privacy-preserving visual reasoning, built into the architecture. Congratulations to Dr. Zehua Cheng (@zehuac) and the FLock Research Team on this well-deserved achievement! We'll share the camera-ready version soon.