Research map / Video / Action recognition
Learning To Align Sequential Actions in the Wild
Check 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 VAVA instead of the strongest listed baseline | COIN phase classification · COIN, labels 0.1 | phase classification accuracy: 35.87 -> 43.77 | Helped |
| Use VAVA instead of the strongest listed baseline | IKEA ASM with background frames · IKEA ASM, background, labels 0.1 | phase classification accuracy: 23.19 -> 29.12 | Helped |
| Use VAVA instead of the strongest listed baseline | Pouring phase classification · Pouring, labels 0.5 | phase classification accuracy: 92.82 -> 91.79 | Hurt |
| Use VAVA instead of the strongest listed baseline | Penn Action phase classification · Penn Action, labels 1.0 | phase classification accuracy: 84.45 -> 84.48 | Helped |
| Add a virtual frame to optimal transport (ablation) | VAVA · IKEA ASM, background, labels 0.1 | phase classification accuracy: 28.03 -> 29.12 | Helped |
| Use a virtual frame rather than thresholding unmatched alignments (ablation) | VAVA · IKEA ASM, background, labels 0.1 | phase classification accuracy: 26.49 -> 29.12 | Helped |
| Add the optimality prior (ablation) | VAVA · IKEA ASM, background, labels 0.1 | phase classification accuracy: 27.03 -> 29.12 | Helped |
| Add the consistency prior (ablation) | VAVA · IKEA ASM, background, labels 0.1 | phase classification accuracy: 24.21 -> 29.12 | Helped |
| Add inter-video contrastive regularization and temporal alignment terms (ablation) | VAVA · IKEA ASM, background, labels 0.1 | phase classification accuracy: 20.38 -> 29.12 | Helped |
| Combine consistency and optimality priors (ablation) | VAVA · IKEA ASM, background, labels 0.1 | phase classification accuracy: 27.03 -> 28.59 | Helped |
| Add KL divergence regularization (ablation) | VAVA · IKEA ASM, background, labels 0.1 | phase classification accuracy: 22.80 -> 24.75 | Helped |
Closest papers by meaning
- Learning by Aligning Videos in Time (CVPR 2021 · similarity 0.91)
- Representation Learning via Global Temporal Alignment and Cycle-Consistency (CVPR 2021 · similarity 0.90)
- Temporal Alignment-Free Video Matching for Few-shot Action Recognition (CVPR 2025 · similarity 0.88)
- Learning to Predict Activity Progress by Self-Supervised Video Alignment (CVPR 2024 · similarity 0.88)
- Efficient and Effective Weakly-Supervised Action Segmentation via Action-Transition-Aware Boundary Alignment (CVPR 2024 · similarity 0.87)
- Align Before Adapt: Leveraging Entity-to-Region Alignments for Generalizable Video Action Recognition (CVPR 2024 · similarity 0.87)
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.