Research map / Video / Video-language understanding
HiTeA: Hierarchical Temporal-Aware Video-Language Pre-training
Qinghao Ye, Guohai Xu, Ming Yan, Haiyang Xu, Qi Qian, Ji Zhang, Fei Huang
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 cross-modal moment exploration (CME) loss (ablation) | HiTeA base objective · MSRVTT retrieval | R@1: 40.0 -> 42.0 | Improved |
| add multi-modal temporal relation exploration (MTRE) loss (ablation) | HiTeA base objective · MSRVTT retrieval | R@1: 40.0 -> 41.6 | Improved |
| combine CME and MTRE (ablation) | HiTeA base objective · MSRVTT retrieval | R@1: 40.0 -> 44.4 | Improved |
| combine CME and MTRE (ablation) | HiTeA base objective · SSv2 Template retrieval | R@1: 80.5 -> 85.6 | Improved |
| combine CME and MTRE (ablation) | HiTeA base objective · MSVD-QA | accuracy: 52.7 -> 55.6 | Improved |
| use CME alone versus MTRE alone (ablation) | HiTeA · MSRVTT retrieval | R@1: 41.6 -> 42.0 | Improved |
| use HiTeA pre-training (5M data) versus prior methods on temporal datasets | HiTeA · SSv2-Template and SSv2-Label retrieval | R@1: 77.0 -> 85.6 (Template); 44.1 -> 55.2 (Label) | Improved |
| scale pre-training corpus from 5M to 17M (ablation) | HiTeA · MSRVTT retrieval | R@1: 44.4 -> 46.8 | Improved |
| shuffle video frames at inference (ablation) | HiTeA on MSRVTT-QA · MSRVTT-QA | accuracy: 45.4 -> 45.2 | No clear effect |
| HiTeA versus LAVENDER on LSMDC-FiB | HiTeA · LSMDC-FiB | accuracy: 56.9 -> 54.6 | Hurt |
Closest papers by meaning
- TOPA: Extending Large Language Models for Video Understanding via Text-Only Pre-Alignment (NeurIPS 2024 · similarity 0.89)
- Towards Balanced Alignment: Modal-Enhanced Semantic Modeling for Video Moment Retrieval (AAAI 2024 · similarity 0.89)
- Momentor: Advancing Video Large Language Model with Fine-Grained Temporal Reasoning (ICML 2024 · similarity 0.88)
- LanguageBind: Extending Video-Language Pretraining to N-modality by Language-based Semantic Alignment (ICLR 2024 · similarity 0.88)
- Video-LaVIT: Unified Video-Language Pre-training with Decoupled Visual-Motional Tokenization (ICML 2024 · similarity 0.88)
- VidLA: Video-Language Alignment at Scale (CVPR 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.
