Research map / Video / Video-language understanding
ETVA: Evaluation of Text-to-Video Alignment via Fine-grained Question Generation and Answering
Kaisi Guan, Zhengfeng Lai, Yuchong Sun, Peng Zhang, Wei Liu, Kieran Liu, Meng Cao, Ruihua Song
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 multi-agent scene-graph question generation instead of vanilla in-context QG (ablation) | ETVA · Human correlation | Spearman's rho: 42.87 -> 58.47 | Improved |
| use multi-agent scene-graph question generation instead of vanilla in-context QG (ablation) | ETVA · Human correlation | Kendall's tau: 35.04 -> 47.16 | Improved |
| add knowledge augmentation (ablation) | ETVA QA · Human annotations | accuracy: 67.34 -> 89.27 | Improved |
| add video understanding stage (ablation) | ETVA QA · Human annotations | accuracy: 82.73 -> 89.27 | Improved |
| add critical reflection stage (ablation) | ETVA QA · Human annotations | accuracy: 68.74 -> 89.27 | Improved |
| add multi-stage reasoning on top of knowledge augmentation alone (ablation) | ETVA QA · Human annotations | Spearman's rho: 33.12 -> 58.47 | Improved |
| use full ETVA QA pipeline instead of direct video LLM answer (ablation) | Qwen2-VL-72B · Human annotations | accuracy: 63.07 -> 89.27 | Improved |
| add knowledge augmentation alone without multi-stage reasoning (ablation) | Direct answer · Human annotations | accuracy: 63.07 -> 65.48 | Improved |
| use ETVA instead of VideoScore as alignment metric | text-to-video alignment metric · Human correlation | Spearman's rho: 31.0 -> 58.5 | Improved |
Closest papers by meaning
- Video-Context Aligned Transformer for Video Question Answering (AAAI 2024 · similarity 0.90)
- Align and Aggregate: Compositional Reasoning with Video Alignment and Answer Aggregation for Video Question-Answering (CVPR 2024 · similarity 0.89)
- LongVideoBench: A Benchmark for Long-context Interleaved Video-Language Understanding (NeurIPS 2024 · similarity 0.89)
- Self-alignment of Large Video Language Models with Refined Regularized Preference Optimization (NeurIPS 2025 · similarity 0.88)
- The Devil is in the Prompts: Retrieval-Augmented Prompt Optimization for Text-to-Video Generation (CVPR 2025 · similarity 0.88)
- Not Another Text Benchmark: Putting the “Visual" Back in Visual Question Answering for Large Video Models (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.
