Research map / Video / Action recognition
Connectionist Temporal Modeling for Weakly Supervised Action Labeling
De-An Huang, Li Fei-Fei, Juan Carlos Niebles
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 k-means clustering visual similarity to CTC (ECTC kmeans) (ablation) | CTC on Breakfast · Breakfast | frame accuracy (%): 21.8 -> 24.5 | Improved |
| add cosine similarity to CTC (ECTC cosine) (ablation) | CTC on Breakfast · Breakfast | frame accuracy (%): 21.8 -> 22.5 | Improved |
| add cosine similarity to CTC (ECTC cosine) (ablation) | CTC on Breakfast · Breakfast | unit accuracy (%): 36.3 -> 36.7 | No clear effect |
| combine k-means and cosine similarities (full ECTC) versus plain CTC (ablation) | CTC on Breakfast · Breakfast | frame accuracy (%): 21.8 -> 27.7 | Improved |
| combine k-means and cosine similarities (full ECTC) versus plain CTC (ablation) | CTC on Breakfast · Breakfast | unit accuracy (%): 36.3 -> 35.6 | No clear effect |
| evaluate all alignments with CTC instead of a single uniform path (ablation) | Uniform baseline on Breakfast · Breakfast | frame accuracy (%): 22.6 -> 21.8 | Hurt |
| evaluate all alignments with CTC instead of a single uniform path (ablation) | Uniform baseline on Breakfast · Breakfast | unit accuracy (%): 33.1 -> 36.3 | Improved |
| use ECTC temporal model instead of OCDC linear classifier at test time | weakly supervised action labeling · Breakfast | frame accuracy (%): 8.9 -> 27.7 | Improved |
Closest papers by meaning
- Two-Stream Networks for Weakly-Supervised Temporal Action Localization With Semantic-Aware Mechanisms (CVPR 2023 · similarity 0.90)
- Weakly Supervised Energy-Based Learning for Action Segmentation (ICCV 2019 · similarity 0.89)
- Cross-Task Relation-Aware Consistency for Weakly Supervised Temporal Action Detection (TPAMI 2025 · similarity 0.89)
- Two-Branch Relational Prototypical Network for Weakly Supervised Temporal Action Localization (TPAMI 2021 · similarity 0.89)
- D3TW: Discriminative Differentiable Dynamic Time Warping for Weakly Supervised Action Alignment and Segmentation (CVPR 2019 · similarity 0.89)
- Similar Modality Enhancement and Action Consistency Learning for Weakly Supervised Temporal Action Localization (AAAI 2025 · similarity 0.88)
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.
