Research map / Video / Video-language understanding
COOT: Cooperative Hierarchical Transformer for Video-Text Representation Learning
Simon Ging, Mohammadreza Zolfaghari, Hamed Pirsiavash, Thomas Brox
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 COOT instead of HSE | ActivityNet-captions video-paragraph retrieval · ActivityNet-captions val1, paragraph→video | R@1: 44.4±0.5 -> 60.8±0.6 | Improved |
| Use COOT instead of HSE | ActivityNet-captions video-paragraph retrieval · ActivityNet-captions val2, paragraph→video | R@1: 32.9 -> 48.5 | Improved |
| Use COOT with HowTo100M-pretrained video-text features instead of MIL-NCE | YouCook2 paragraph-to-video retrieval · YouCook2, paragraph→video | R@1: 61.9* -> 77.2±1.0 | Improved |
| Add attention-aware feature aggregation rather than max pooling (ablation) | COOT · ActivityNet-captions val1, paragraph→video | R@1: 58.2±0.5 -> 59.0±0.5 | Improved |
| Add cross-modal cycle-consistency loss (ablation) | COOT with AFA and CoT · ActivityNet-captions val1, paragraph→video | R@1: 59.8±1.1 -> 60.8±0.6 | Improved |
| Add the contextual transformer (ablation) | COOT with AFA and CMC · ActivityNet-captions val1, paragraph→video | R@1: 59.5±0.5 -> 60.8±0.6 | Improved |
| Replace BERT with GloVe text features (ablation) | COOT · ActivityNet-captions val1, paragraph→video | R@1: 60.8±0.6 -> 56.5±1.1 | Hurt |
| Remove the low-level alignment loss (ablation) | COOT full objective · ActivityNet-captions val1, paragraph→video | R@1: 60.8±0.6 -> 58.5±1.1 | Hurt |
| Add global-context clustering loss (ablation) | COOT · ActivityNet-captions val1, paragraph→video | R@1: 60.8±0.6 -> 60.8±0.6 | No clear effect |
| Use attention-aware feature aggregation instead of average pooling at the high level (ablation) | COOT · ActivityNet-captions val1, paragraph→video | R@1: 60.8±0.6 -> 42.6±0.4 | Hurt |
| Use two high-level attention-aware feature aggregation modules instead of one (ablation) | COOT · ActivityNet-captions val1, paragraph→video | R@1: 42.6±0.4 -> 42.8±0.1 | No clear effect |
| Use HowTo100M-pretrained features instead of classification-pretrained features (ablation) | COOT on YouCook2 · YouCook2, paragraph→video | R@1: 50.4±2.6 -> 77.2±1.0 | Improved |
Later papers that compared with this method
1 later accepted paper reports COOT as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- Text With Knowledge Graph Augmented Transformer for Video Captioning (CVPR 2023 · better on 1 of 1 shared result)
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 0 of 1 on most shared results.
- ActBERT — ActBERT: Learning Global-Local Video-Text Representations (CVPR 2020) · better on 0 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 Joint Modelling Based on Hierarchical Transformer for Co-summarization (TPAMI 2022 · similarity 0.87)
- SViTT: Temporal Learning of Sparse Video-Text Transformers (CVPR 2023 · similarity 0.87)
- Weakly Supervised Video Representation Learning With Unaligned Text for Sequential Videos (CVPR 2023 · similarity 0.86)
- CogVideoX: Text-to-Video Diffusion Models with An Expert Transformer (ICLR 2025 · similarity 0.85)
- TOPA: Extending Large Language Models for Video Understanding via Text-Only Pre-Alignment (NeurIPS 2024 · similarity 0.85)
- HierarQ: Task-Aware Hierarchical Q-Former for Enhanced Video Understanding (CVPR 2025 · similarity 0.84)
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.
