[2603.27817] Towards Context-Aware Image Anonymization with Multi-Agent Reasoning

Computer Science > Computer Vision and Pattern Recognition arXiv:2603.27817 (cs) [Submitted on 29 Mar 2026] Title:Towards Context-Aware Image Anonymization with Multi-Agent Reasoning Authors:Robert Aufschläger, Jakob Folz, Gautam Savaliya, Manjitha D Vidanalage, Michael Heigl, Martin Schramm View a PDF of the paper titled Towards Context-Aware Image Anonymization with Multi-Agent Reasoning, by Robert Aufschl\"ager and 5 other authors View PDF HTML (experimental) Abstract:Street-level imagery contains personally identifiable information (PII), some of which is context-dependent. Existing anonymization methods either over-process images or miss subtle identifiers, while API-based solutions compromise data sovereignty. We present an agentic framework CAIAMAR (\underline{C}ontext-\underline{A}ware \underline{I}mage \underline{A}nonymization with \underline{M}ulti-\underline{A}gent \underline{R}easoning) for context-aware PII segmentation with diffusion-based anonymization, combining pre-defined processing for high-confidence cases with multi-agent reasoning for indirect identifiers. Three specialized agents coordinate via round-robin speaker selection in a Plan-Do-Check-Act (PDCA) cycle, enabling large vision-language models to classify PII based on spatial context (private vs. public property) rather than rigid category rules. The agents implement spatially-filtered coarse-to-fine detection where a scout-and-zoom strategy identifies candidates, open-vocabulary segmentation pr...