ARGOS: Who, Where, and When in Agentic Multi-Camera Person Search

2026-04-14

AI Summary

This work addresses the challenge of person retrieval in multi-camera settings based on vague eyewitness descriptions by formulating it for the first time as an interactive reasoning task for an embodied agent, which must locate the target under conditions of information asymmetry and a limited number of interaction rounds. The proposed approach centers on a large language model acting as the core reasoning agent, augmented with a spatio-temporal topology graph (STTG) that encodes camera topology and empirically derived transition times, along with domain-specific tools to enable coordinated semantic, spatial, and temporal reasoning. Evaluated across 2,691 tasks in 14 real-world scenarios, the best-performing model achieves TWS scores of 0.383 and 0.590 on Track 2 and Track 3, respectively. Ablation studies further demonstrate that removing domain-specific tools can reduce accuracy by up to 49.6 percentage points.

Technology Category

• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning
• Computer Vision: Visual Reasoning & Symbolic Representations
• Planning, Routing, and Scheduling: Model-Based Reasoning

Application Category

• Search and Retrieval-Augmented AI: Agentic search
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

We introduce ARGOS, the first benchmark and framework that reformulates multi-camera person search as an interactive reasoning problem requiring an agent to plan, question, and eliminate candidates under information asymmetry. An ARGOS agent receives a vague witness statement and must decide what to ask, when to invoke spatial or temporal tools, and how to interpret ambiguous responses, all within a limited turn budget. Reasoning is grounded in a Spatio-Temporal Topology Graph (STTG) encoding camera connectivity and empirically validated transition times. The benchmark comprises 2,691 tasks across 14 real-world scenarios in three progressive tracks: semantic perception (Who), spatial reasoning (Where), and temporal reasoning (When). Experiments with four LLM backbones show the benchmark is far from solved (best TWS: 0.383 on Track 2, 0.590 on Track 3), and ablations confirm that removing domain-specific tools drops accuracy by up to 49.6 percentage points.

Problem

• multi-camera person search
• interactive reasoning
• information asymmetry
• spatio-temporal reasoning
• agent-based search

Innovation

• agentic reasoning
• multi-camera person search
• spatio-temporal topology graph
• interactive questioning
• information asymmetry