Research map / Video / Video-language understanding
A Multi-Agent Perception-Action Alliance for Efficient Long Video Reasoning
Yichang Xu, Gaowen Liu, Ramana Rao Kompella, Tiansheng Huang, Sihao Hu, Fatih Ilhan, Selim Furkan Tekin, Zachary Yahn, Ling Liu
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 |
|---|---|---|---|
| sample perception frames randomly instead of per event block (ablation) | A4VL · EgoSchema (test, 500 QA) | accuracy: 79.6 -> 82.2 | Improved |
| use event-based blocks instead of uniform blocks for action frames (ablation) | A4VL · EgoSchema (test, 500 QA) | accuracy: 80.2 -> 82.2 | Improved |
| require full consensus instead of majority consensus (ablation) | A4VL · EgoSchema (test, 500 QA) | accuracy: 81.4 -> 82.2 | Improved |
| require full consensus instead of majority consensus (ablation) | A4VL · EgoSchema (test, 500 QA) latency | seconds per sample: 26s -> 37s | Hurt |
| add agent pruning (vs sum selection) (ablation) | A4VL · EgoSchema (test, 500 QA) | accuracy: 80.8 -> 82.2 | Improved |
| add agent pruning (vs majority selection) (ablation) | A4VL · EgoSchema (test, 500 QA) | accuracy: 79.4 -> 82.2 | Improved |
| add agent pruning (ablation) | A4VL · EgoSchema (test, 500 QA) latency | seconds per sample: 60s -> 37s | Improved |
| allow more collaboration rounds (ablation) | A4VL · all five benchmarks | accuracy | Improved |
| replace single strong MLLM with A4VL multi-agent framework | InternVL3-78B · EgoSchema (test, 500 QA) | accuracy: 76.8 -> 82.2 | Improved |
| use A4VL instead of GPT-4o on MLVU | GPT-4o · MLVU latency | seconds per sample: 127s -> 74s | Improved |
| use A4VL instead of GPT-4o | GPT-4o · LongVideoBench | accuracy: 66.7 -> 72.2 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- LVAgent — LVAgent: Long Video Understanding by Multi-Round Dynamical Collaboration of MLLM Agents (ICCV 2025) · better on 5 of 5 shared results
- VideoAgent — VideoAgent: A Memory-augmented Multimodal Agent for Video Understanding (ECCV 2024) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Think, Then Verify: A Hypothesis-Verification Multi-Agent Framework for Long Video Understanding (CVPR 2026 · similarity 0.91)
- ReAgent-V: A Reward-Driven Multi-Agent Framework for Video Understanding (NeurIPS 2025 · similarity 0.90)
- Enhancing Video-LLM Reasoning via Agent-of-Thoughts Distillation (CVPR 2025 · similarity 0.89)
- ReWatch-R1: Boosting Complex Video Reasoning in Large Vision-Language Models through Agentic Data Synthesis (ICLR 2026 · similarity 0.89)
- Agent-X: Evaluating Deep Multimodal Reasoning in Vision-Centric Agentic Tasks (ICLR 2026 · similarity 0.88)
- Act2See: Emergent Active Visual Perception for Video Reasoning (CVPR 2026 · similarity 0.88)
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.
