Research map / Video / Video-language understanding
VideoAds for Fast-Paced Video Understanding
Zheyuan Zhang, Wanying Dou, Linkai Peng, Hongyi Pan, Ulas Bagci, Boqing Gong
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 |
|---|---|---|---|
| add speech audio transcript (Whisper) to frame input (ablation) | LongVA · VideoAds | overall accuracy (%): 42.33 -> 48.18 | Improved |
| add speech audio transcript to frame input (ablation) | LLaVA-Video-72B · VideoAds | overall accuracy (%): 66.58 -> 71.73 | Improved |
| add speech audio transcript to frame input (ablation) | Qwen2.5-VL-72B · VideoAds | overall accuracy (%): 64.33 -> 67.48 | Improved |
| add speech audio transcript to frame input (ablation) | GPT-4o · VideoAds | overall accuracy (%): 65.86 -> 66.47 | Mixed |
| increase sampled frames from 32 to 128 (ablation) | Qwen2.5-VL-72B · VideoAds | overall accuracy (%): 64.33 -> 73.35 | Improved |
| increase sampled frames from 32 to 128 (ablation) | LLaVA-Video-72B · VideoAds | overall accuracy (%): 66.58 -> 68.45 | Mixed |
| increase sampled frames from 32 to 128 (ablation) | GPT-4o · VideoAds | overall accuracy (%): 65.57 -> 66.82 | Mixed |
| add guided chain-of-thought intermediate questions (ablation) | GPT-4o · VideoAds reasoning | reasoning accuracy (%): 61 -> 70 | Improved |
| add guided chain-of-thought intermediate questions (ablation) | LongVA-7B, LLaVA-NeXT-32B, LLaVA-Video-72B · VideoAds reasoning | reasoning accuracy | Hurt |
| add guided chain-of-thought intermediate questions (ablation) | Qwen2.5-VL · VideoAds reasoning | reasoning accuracy | Improved |
| remove visual input (text-only GPT-4o) (ablation) | GPT-4o · VideoAds | overall accuracy (%): 66.82 -> 21.27 | Hurt |
Closest papers by meaning
- MMBench-Video: A Long-Form Multi-Shot Benchmark for Holistic Video Understanding (NeurIPS 2024 · similarity 0.89)
- Video-MME: The First-Ever Comprehensive Evaluation Benchmark of Multi-modal LLMs in Video Analysis (CVPR 2025 · similarity 0.89)
- MotionBench: Benchmarking and Improving Fine-grained Video Motion Understanding for Vision Language Models (CVPR 2025 · similarity 0.88)
- Needle In A Video Haystack: A Scalable Synthetic Evaluator for Video MLLMs (ICLR 2025 · similarity 0.88)
- MVBench: A Comprehensive Multi-modal Video Understanding Benchmark (CVPR 2024 · similarity 0.87)
- MMOU: A Massive Multi-Task Omni Understanding and Reasoning Benchmark for Long and Complex Real-World Videos (NeurIPS 2026 · similarity 0.87)
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.
