Research map / Large language models / Retrieval and knowledge
ARGOS: Who, Where, and When in Agentic Multi-Camera Person Search
Myungchul Kim, Kwanyong Park, Junmo Kim, In Kweon
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use structured tool-calling instead of direct LLM inference | GPT-4o agent · ARGOS Track 1 (Who) | SR@1 (%): 73.3 -> 81.1 | Improved |
| use structured tool-calling instead of direct LLM inference | GPT-4o agent · ARGOS Track 1 (Who) | Parsing accuracy (%): 93.0 -> 90.8 | Hurt |
| add temporal tool (check_temporal) (ablation) | ARGOS Agent (GPT-4o) · ARGOS Track 3 (When) | Top-1 accuracy (%): 31.0 -> 80.6 | Improved |
| add spatial tool (ablation) | ARGOS Agent (GPT-4o) · ARGOS Track 2 (Where) | Top-1 accuracy (%): 40.7 -> 74.5 | Improved |
| add all tools (ablation) | ARGOS Agent (GPT-4o) · ARGOS Track 3 (When) | Top-1 accuracy (%): 11.4 -> 80.6 | Improved |
| add strategic LLM reasoning (ablation) | ARGOS Agent (GPT-4o) · ARGOS Track 3 (When) | TWS: 0.373 -> 0.567 | Improved |
| add strategic LLM reasoning (ablation) | ARGOS Agent (GPT-4o) · ARGOS Track 3 (When) | Top-1 accuracy (%): 76.9 -> 80.6 | Improved |
| add strategic LLM reasoning (ablation) | ARGOS Agent (GPT-4o) · ARGOS Track 2 (Where) | TWS: 0.136 -> 0.323 | Improved |
| add strategic LLM reasoning (ablation) | ARGOS Agent (GPT-4o) · ARGOS Track 2 (Where) | Top-1 accuracy (%): 47.6 -> 74.5 | Improved |
| let agent invoke check_temporal at will instead of temporal-first (ablation) | ARGOS Agent (GPT-4o) · ARGOS Track 3 (When) | Top-1 accuracy (%): 80.6 -> 83.9 | Improved |
| let agent invoke check_temporal at will instead of temporal-first (ablation) | ARGOS Agent (GPT-4o) · ARGOS Track 3 (When) | TWS: 0.567 -> 0.572 | No clear effect |
| add dialogue interaction on top of single-pass inference (Full vs Initial) (ablation) | single-pass baseline · ARGOS Track 2 (Where) | Top-1 accuracy (%): 64.7 -> 56.2 | Hurt |
Closest papers by meaning
- Agent-X: Evaluating Deep Multimodal Reasoning in Vision-Centric Agentic Tasks (ICLR 2026 · similarity 0.86)
- ARK: A Dual-Axis Multimodal Retrieval Benchmark along Reasoning and Knowledge (NeurIPS 2026 · similarity 0.86)
- REVERSE: Reinforcing Evidence Verification and Search for Agentic Image Geolocation (NeurIPS 2026 · similarity 0.86)
- SMMBench: A Benchmark for Source-Distributed Multimodal Agent Memory (NeurIPS 2026 · similarity 0.86)
- HumanPCR: Probing MLLM Capabilities in Diverse Human-Centric Scenes (ICLR 2026 · similarity 0.85)
- Navigating Massive Visual Context in Retrieval-Augmented Generation via Multimodal Memory Graph (ICML 2026 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
