Research map / Vision-language and multimodal / Cross-modal learning
Asynchronous Matching with Dynamic Sampling for Multimodal Dataset Distillation
Ding Qi, Jian Li, Shuguang Dou, Zifan Song, Junyao Gao, Yabiao Wang, Chengjie Wang, Cairong Zhao
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 |
|---|---|---|---|
| replace synchronous with asynchronous trajectory matching plus prototype init (AMD) | LoRS · Flickr30k 200 pairs | IR@1: 8.6 -> 13.1 | Improved |
| AMD vs LoRS | LoRS · Flickr30k 200 pairs | TR@10: 53.4 -> 56.2 | Improved |
| AMD vs LoRS | LoRS · COCO 200 pairs | IR@1: 2.4 -> 3.8 | Improved |
| increase pair budget from 99 to 499 | AMD vs LoRS · Flickr30k | IR@1 gain over LoRS: +2.1 (99) -> +5.8 (499) | Improved |
| add asynchronous trajectory matching to LoRS baseline (ablation) | LoRS · Flickr30k 200 pairs | IR@1: 8.6 -> 12.1 | Improved |
| add semantics-aware prototype mining (SPM) to LoRS baseline (ablation) | LoRS · Flickr30k 200 pairs | IR@1: 8.6 -> 9.1 | Improved |
| add SPM on top of asynchronous matching (ablation) | AMD · Flickr30k 200 pairs | IR@1: 12.1 -> 13.1 | Improved |
| use AMD distilled data to train other image encoders | AMD vs LoRS · Flickr30k 499 pairs, ResNet+BERT | IR@1: 3.3 -> 4.1 | Improved |
| distill to 10% synthetic data instead of full data | AMD (NFNet+CLIP) · Flickr30k | IR@1: 49.8 -> 47.9 | Mixed |
Closest papers by meaning
- Beyond Modality Collapse: Representation Blending for Multimodal Dataset Distillation (NeurIPS 2025 · similarity 0.88)
- CovMatch: Cross-Covariance Guided Multimodal Dataset Distillation with Trainable Text Encoder (NeurIPS 2025 · similarity 0.88)
- Efficient Multimodal Dataset Distillation via Generative Models (NeurIPS 2025 · similarity 0.87)
- Towards Stable and Storage-efficient Dataset Distillation: Matching Convexified Trajectory (CVPR 2025 · similarity 0.87)
- Continuous-Time Distribution Matching for Few-Step Diffusion Distillation (NeurIPS 2026 · similarity 0.86)
- Low-Rank Similarity Mining for Multimodal Dataset Distillation (ICML 2024 · similarity 0.86)
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.
