Research map / Video / Action recognition
TA2N: Two-Stage Action Alignment Network for Few-Shot Action Recognition
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 two-stage action alignment (TA²N) instead of OTAM* | ResNet-50 few-shot action recognition · HMDB51 5-way 1-shot | accuracy: 54.5 -> 59.7 | Helped |
| use two-stage action alignment (TA²N) instead of ProtoNet | ResNet-50 few-shot action recognition · UCF101 5-way 5-shot | accuracy: 89.6 -> 95.1 | Helped |
| use two-stage action alignment (TA²N) instead of OTAM* | ResNet-50 few-shot action recognition · SSv2 5-way 5-shot | accuracy: 52.3 -> 61.0 | Helped |
| use two-stage action alignment (TA²N) instead of OTAM* | ResNet-50 few-shot action recognition · Kinetics-CMN 5-way 1-shot | accuracy: 73.0 -> 72.8 | Hurt |
| add TTM to the ProtoNet baseline (ablation) | Few-shot action recognition · UCF101 5-way 1-shot | accuracy: 74.0 -> 78.5 | Helped |
| add TTM and TC to the baseline (ablation) | Few-shot action recognition · SSv2 5-way 1-shot | accuracy: 40.1 -> 46.3 | Helped |
| add SC to TTM and TC (ablation) | Few-shot action recognition · UCF101 5-way 1-shot | accuracy: 80.9 -> 81.9 | Helped |
| add SC to TTM and TC (ablation) | Few-shot action recognition · SSv2 5-way 1-shot | accuracy: 46.3 -> 47.6 | Helped |
| add TTM before TC and SC (ablation) | Few-shot action recognition · UCF101 5-way 1-shot | accuracy: 81.0 -> 81.9 | Helped |
| add TTM before TC and SC (ablation) | Few-shot action recognition · SSv2 5-way 1-shot | accuracy: 47.0 -> 47.6 | Helped |
| replace enumeration-based SC with mask-based SC (ablation) | TA²N spatial coordination · HMDB51 5-way 1-shot | accuracy: 57.6 -> 59.9 | Helped |
| replace grid-based SC with mask-based SC (ablation) | TA²N spatial coordination · UCF101 5-way 1-shot | accuracy: 81.4 -> 81.9 | Helped |
Closest papers by meaning
- Temporal Alignment-Free Video Matching for Few-shot Action Recognition (CVPR 2025 · similarity 0.91)
- Searching for Better Spatio-temporal Alignment in Few-Shot Action Recognition (NeurIPS 2022 · similarity 0.89)
- Temporal-Relational CrossTransformers for Few-Shot Action Recognition (CVPR 2021 · similarity 0.88)
- Frame Order Matters: A Temporal Sequence-Aware Model for Few-Shot Action Recognition (AAAI 2025 · similarity 0.87)
- Storyboard-guided Alignment for Fine-grained Video Action Recognition (NeurIPS 2025 · similarity 0.86)
- Learning Causal Domain-Invariant Temporal Dynamics for Few-Shot Action Recognition (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.