Research map / Video / Video-language understanding
A Dataset and Exploration of Models for Understanding Video Data Through Fill-In-The-Blank Question-Answering
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 static 2D and moving 3D visual features to question text | MovieFIB fill-in-the-blank model · MovieFIB test | accuracy: 34.4 -> 35.7 | Helped |
| Combine 2D and 3D visual features rather than using either alone (ablation) | MovieFIB fill-in-the-blank model · MovieFIB test | accuracy: 34.9 -> 35.7; 34.5 -> 35.7 | Helped |
| Initialize the text encoder from the Text-only baseline (ablation) | GoogLeNet-2D MovieFIB model · MovieFIB test | accuracy: 34.9 -> 35.3 | Helped |
| Reduce input and output vocabulary sizes (ablation) | Text-only MovieFIB model · MovieFIB test | accuracy: 34.4 -> 35.0 | Helped |
| Reduce input and output vocabulary sizes (ablation) | GoogLeNet-2D + C3D MovieFIB model · MovieFIB test | accuracy: 35.7 -> 36.3 | Helped |
| Provide video along with the question | Human answerers · MovieFIB test subset | accuracy: 30.2 -> 68.7 | Helped |
| Increase training data from 10% to more of the training set | GoogLeNet-2D + C3D MovieFIB model · MovieFIB | accuracy | Helped |
| Increase training data from 10% to more of the training set | Text-only MovieFIB model · MovieFIB | accuracy | Helped |
| Use only 2 sampled video frames instead of the default 25 (ablation) | GoogLeNet-2D MovieFIB model · MovieFIB validation | accuracy | No clear effect |
Closest papers by meaning
- PerceptionLM: Open-Access Data and Models for Detailed Visual Understanding (NeurIPS 2025 · similarity 0.90)
- Not Another Text Benchmark: Putting the “Visual" Back in Visual Question Answering for Large Video Models (NeurIPS 2026 · similarity 0.89)
- LongVideoBench: A Benchmark for Long-context Interleaved Video-Language Understanding (NeurIPS 2024 · similarity 0.89)
- HourVideo: 1-Hour Video-Language Understanding (NeurIPS 2024 · similarity 0.89)
- Distilling Vision-Language Models on Millions of Videos (CVPR 2024 · similarity 0.88)
- MMBench-Video: A Long-Form Multi-Shot Benchmark for Holistic Video Understanding (NeurIPS 2024 · 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.