Research map / Video / Video-language understanding
Just Ask: Learning To Answer Questions From Millions of Narrated Videos
Antoine Yang, Antoine Miech, Josef Sivic, Ivan Laptev, Cordelia Schmid
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 |
|---|---|---|---|
| pretrain on HowToVQA69M before finetuning | VQA-T · iVQA (finetuned) | top-1 accuracy: 23.0 -> 35.4 | Improved |
| pretrain on HowToVQA69M instead of HowTo100M | VQA-T · iVQA (finetuned) | top-1 accuracy: 28.1 -> 35.4 | Improved |
| pretrain on HowToVQA69M before finetuning | VQA-T · MSRVTT-QA (finetuned) | top-1 accuracy: 39.6 -> 41.5 | Improved |
| pretrain on HowToVQA69M before finetuning | VQA-T · How2QA (finetuned) | top-1 accuracy: 80.8 -> 84.4 | Improved |
| add video input (VQA-T instead of language-only QA-T) | VQA-T trained on HowToVQA69M · iVQA (zero-shot) | top-1 accuracy: 4.4 -> 12.2 | Improved |
| train on HowToVQA69M instead of HowTo100M | VQA-T · iVQA (zero-shot) | top-1 accuracy: 1.9 -> 12.2 | Improved |
| remove duplicate negative answers in contrastive loss (ablation) | VQA-T · iVQA (zero-shot) | top-1 accuracy: 10.9 -> 12.2 | Improved |
| add MLM loss in pretraining (ablation) | VQA-T · iVQA (finetuned) | top-1 accuracy: 34.3 -> 35.4 | Improved |
| add MLM loss in pretraining (ablation) | VQA-T · iVQA (zero-shot) | top-1 accuracy: 12.1 -> 12.2 | Improved |
| increase pretraining data from 1% to 100% of HowToVQA69M (ablation) | VQA-T · iVQA (finetuned) | top-1 accuracy: 24.2 -> 35.4 | Improved |
| use transformer QA generation instead of rule-based Heilman et al. (ablation) | VQA-T · iVQA (finetuned) | top-1 accuracy: 31.4 -> 35.4 | Improved |
| use transformer QA generation instead of rule-based Heilman et al. (ablation) | VQA-T · iVQA (zero-shot) | top-1 accuracy: 7.4 -> 12.2 | Improved |
Closest papers by meaning
- Learning to Answer Visual Questions From Web Videos (TPAMI 2022 · similarity 0.98)
- Leveraging Video Descriptions to Learn Video Question Answering (AAAI 2017 · similarity 0.92)
- Zero-Shot Video Question Answering via Frozen Bidirectional Language Models (NeurIPS 2022 · similarity 0.91)
- Open-vocabulary Video Question Answering: A New Benchmark for Evaluating the Generalizability of Video Question Answering Models (ICCV 2023 · similarity 0.90)
- Knowledge-Constrained Answer Generation for Open-Ended Video Question Answering (AAAI 2023 · similarity 0.90)
- Video-Context Aligned Transformer for Video Question Answering (AAAI 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.
