Research map / Video / Action recognition
Hybrid Relation Guided Set Matching for Few-Shot Action Recognition
Xiang Wang, Shiwei Zhang, Zhiwu Qing, Mingqian Tang, Zhengrong Zuo, Changxin Gao, Rong Jin, Nong Sang
Check 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 HyRSM instead of the strongest listed baseline | few-shot action recognition · SSv2-Full 1-shot | accuracy: 49.2 -> 54.3 | Improved |
| use HyRSM instead of the strongest listed baseline | few-shot action recognition · UCF101 5-shot | accuracy: 96.1 -> 94.7 | Hurt |
| add intra-relation modeling (ablation) | ProtoNet on SSv2-Full · SSv2-Full 1-shot | accuracy: 35.2 -> 41.2 | Improved |
| add inter-relation modeling (ablation) | ProtoNet on SSv2-Full · SSv2-Full 1-shot | accuracy: 35.2 -> 43.7 | Improved |
| replace baseline distance with Bi-MHM (ablation) | ProtoNet on SSv2-Full · SSv2-Full 1-shot | accuracy: 35.2 -> 44.6 | Improved |
| combine intra-relation, inter-relation, and Bi-MHM (ablation) | ProtoNet on SSv2-Full · SSv2-Full 1-shot | accuracy: 35.2 -> 54.3 | Improved |
| add intra-relation and inter-relation modeling (ablation) | OTAM on SSv2-Full · SSv2-Full 1-shot | accuracy: 42.8 -> 51.7 | Improved |
| use support videos only, excluding the query video from relation modeling (ablation) | HyRSM on SSv2-Full · SSv2-Full 1-shot | accuracy: 52.1 -> 54.3 | Hurt |
| use support videos only, excluding the query video from relation modeling (ablation) | HyRSM on Kinetics · Kinetics 1-shot | accuracy: 73.4 -> 73.7 | Hurt |
| use bidirectional Bi-MHM instead of bidirectional OTAM (ablation) | few-shot action recognition on SSv2-Full · SSv2-Full 1-shot | accuracy: 42.8 -> 44.6 | Improved |
| increase the number of sampled frames (ablation) | HyRSM · SSv2-Full 5-way 1-shot | accuracy | Improved |
Closest papers by meaning
- Boosting Few-shot Action Recognition with Graph-guided Hybrid Matching (ICCV 2023 · similarity 0.90)
- Task-Specific Distance Correlation Matching for Few-Shot Action Recognition (AAAI 2026 · similarity 0.89)
- Few-Shot Action Recognition with Hierarchical Matching and Contrastive Learning (ECCV 2022 · similarity 0.89)
- Temporal Alignment-Free Video Matching for Few-shot Action Recognition (CVPR 2025 · similarity 0.88)
- TA2N: Two-Stage Action Alignment Network for Few-Shot Action Recognition (AAAI 2022 · similarity 0.88)
- Frame Order Matters: A Temporal Sequence-Aware Model for Few-Shot Action Recognition (AAAI 2025 · similarity 0.85)
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.
