Research map / Video / Video-language understanding
Appearance-Motion Decomposed Alignment for Text-Video Retrieval
Meng Meng, Zichang Tan, Yong Zhang, Xu Zhou
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 all DAL and MML modules (full AMD-Net vs plain baseline) (ablation) | AMD-Net (CLIP ViT-B/32) · MSRVTT | R@1: 46.2 -> 55.2 | Improved |
| add SVD-based prototype initialization instead of random (ablation) | AMD-Net (CLIP ViT-B/32) · MSRVTT | R@1: 49.1 -> 55.2 | Improved |
| add SVD-based prototype initialization instead of learnable (ablation) | AMD-Net (CLIP ViT-B/32) · MSRVTT | R@1: 52.8 -> 55.2 | Improved |
| add SVD-based initialization instead of learnable (ablation) | AMD-Net (CLIP ViT-B/32) · MSRVTT | MnR: 11.6 -> 9.5 lower is better; improves | Improved |
| add high-order similarity over first-order cross-aggregation in DAL (ablation) | AMD-Net (CLIP ViT-B/32) · MSRVTT | R@1 | Improved |
| add slow motion learning to fast motion only (ablation) | AMD-Net (CLIP ViT-B/32) · MSRVTT | R@1: 51.2 -> 55.2 | Improved |
| add fast motion learning to single-motion variant (ablation) | AMD-Net (CLIP ViT-B/32) · MSRVTT | R@1: 52.9 -> 55.2 | Improved |
| increase prototype number I from 1 to 10 (ablation) | AMD-Net (CLIP ViT-B/32) · MSRVTT | R@1: 54.5 -> 55.2 | Improved |
| increase prototype number I from 10 to 20 (ablation) | AMD-Net (CLIP ViT-B/32) · MSRVTT | R@1: 55.2 -> 55.0 | No clear effect |
| use CLIP ViT-B/16 instead of ViT-B/32 (ablation) | AMD-Net (CLIP ViT-B/32) · MSRVTT | R@1: 55.2 -> 59.0 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- Cap4Video* — Cap4Video: What Can Auxiliary Captions Do for Text-Video Retrieval? (CVPR 2023) · better on 1 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
- Tri-Modal Motion Retrieval by Learning a Joint Embedding Space (CVPR 2024 · similarity 0.87)
- Enhanced Motion-Text Alignment for Image-to-Video Transfer Learning (CVPR 2024 · similarity 0.85)
- Don't Judge by the Look: Towards Motion Coherent Video Representation (ICLR 2024 · similarity 0.85)
- Mitigating Semantic Collapse in Partially Relevant Video Retrieval (NeurIPS 2025 · similarity 0.85)
- Towards Efficient and Effective Text-to-Video Retrieval with Coarse-to-Fine Visual Representation Learning (AAAI 2024 · similarity 0.85)
- DiscoVLA: Discrepancy Reduction in Vision, Language, and Alignment for Parameter-Efficient Video-Text Retrieval (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.
