Research map / Video / Video-language understanding
Progressive Online Video Understanding with Evidence-Aligned Timing and Transparent Decisions
Kecheng Zhang, Zongxin Yang, Mingfei Han, Haihong Hao, Yunzhi Zhuge, Changlin Li, junhan zhao, Zhihui Li, Xiaojun Chang
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 |
|---|---|---|---|
| add HPSI multi-level aggregation instead of single-shot adaptive pooling before the LLM (ablation) | Thinking-QwenVL · OVOBench | overall accuracy: 43.4 -> 46.9 (-3.5 when pooled) | Improved |
| add HPSI multi-level aggregation instead of single-shot adaptive pooling before the LLM (ablation) | Thinking-QwenVL · VideoMME | overall accuracy: 48.9 -> 56.3 (-7.4 when pooled) | Improved |
| add the first-frame token (ablation) | Thinking-QwenVL · OVOBench | overall accuracy: 42.6 -> 46.9 | Improved |
| add level-3 aggregation tokens (ablation) | Thinking-QwenVL · OVOBench | overall accuracy: 46.0 -> 46.9 on OVOBench; 56.0 -> 56.3 on VideoMME | Mixed |
| add level-2 and level-3 aggregation tokens (ablation) | Thinking-QwenVL · OVOBench | overall accuracy: 49.6 -> 46.9 on OVOBench; 54.7 -> 56.3 on VideoMME | Mixed |
| add ATDM decision making to Flash-VStream | Flash-VStream · StreamingBench | accuracy: 22.53 -> 26.58 | Improved |
| replace graded progress/confidence (rho, c) with a binary answerable flag (ablation) | Thinking-QwenVL · online benchmarks | accuracy: 3.62% accuracy drop when removed | Improved |
| add question decomposition (P2) (ablation) | Flash-VStream · online benchmarks | accuracy: 3.60% accuracy drop when removed | Improved |
| add streaming captioning (P3) (ablation) | Flash-VStream · online benchmarks | accuracy: 4.49% accuracy drop when removed | Improved |
| use 3D RoPE instead of offset sequential positional embedding (OSPR) (ablation) | Thinking-QwenVL · OVOBench | overall accuracy: 43.3 -> 46.9 | Improved |
| use full-frame (100%) attention redistribution rather than 93.75% frame reduction | Thinking-QwenVL · OVOBench | overall accuracy: 46.9 -> 52.5 | Improved |
| apply HPSI and ATDM to Qwen2.5-VL baseline | Qwen2.5-VL-7B · RTVBench | overall accuracy: 32.75 -> 35.87 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 4 of 4 on most shared results.
- Flash-VStream — Flash-VStream: Efficient Real-Time Understanding for Long Video Streams (ICCV 2025) · better on 3 of 3 shared results
- Dispider — Dispider: Enabling Video LLMs with Active Real-Time Interaction via Disentangled Perception, Decision, and Reaction (CVPR 2025) · better on 2 of 2 shared results
- MovieChat — MovieChat: From Dense Token to Sparse Memory for Long Video Understanding (CVPR 2024) · better on 1 of 1 shared result
- VideoLLM-online — VideoLLM-online: Online Video Large Language Model for Streaming Video (CVPR 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
- Video Streaming Thinking: VideoLLMs Can Watch and Think Simultaneously (ECCV 2026 · similarity 0.91)
- Thinking in Streaming Video (ECCV 2026 · similarity 0.91)
- Select Less, Reason More: Prioritizing Evidence Purity for Video Reasoning (CVPR 2026 · similarity 0.90)
- When Thinking Drifts: Evidential Grounding for Robust Video Reasoning (NeurIPS 2025 · similarity 0.89)
- Think-as-You-See: Streaming Chain-of-Thought Reasoning for Large Vision-Language Models (CVPR 2026 · similarity 0.89)
- What Should a Streaming Video Model Remember? (NeurIPS 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.
