Research map / Video / Action recognition
End-to-End Semi-Supervised Learning for Video Action Detection
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 semi-supervised spatio-temporal consistency with limited labeled data | I3D action detector · UCF101-24 | v-mAP@0.5: 67.5 -> 72.1 | Helped |
| Use semi-supervised spatio-temporal consistency with limited labeled data | I3D action detector · JHMDB-21 | v-mAP@0.5: 58.5 -> 63.5 | Helped |
| Add classification consistency to supervised training (ablation) | I3D action detector · UCF101-24 | f-mAP@0.5: 61.0 -> 65.0 | Helped |
| Add spatio-temporal localization consistency to supervised training (ablation) | I3D action detector · UCF101-24 | f-mAP@0.5: 61.0 -> 69.8 | Helped |
| Add classification consistency alongside localization consistency (ablation) | I3D action detector · UCF101-24 | f-mAP@0.5: 69.8 -> 69.5 | No clear effect |
| Replace variance attention with cyclic variance (ablation) | Temporal coherence localization consistency · UCF101-24 | v-mAP@0.5: 70.3 -> 70.8 | Helped |
| Add non-attentive L2 loss to cyclic variance (ablation) | Temporal coherence localization consistency · UCF101-24 | v-mAP@0.5: 70.8 -> 72.1 | Helped |
| Add non-attentive L2 loss to cyclic variance (ablation) | Temporal coherence localization consistency · JHMDB-21 | v-mAP@0.5: 61.5 -> 63.5 | Helped |
| Add non-attentive L2 loss to gradient smoothness (ablation) | Gradient smoothness localization consistency · UCF101-24 | v-mAP@0.5: 72.4 -> 72.0 | Hurt |
| Add non-attentive L2 loss to gradient smoothness (ablation) | Gradient smoothness localization consistency · JHMDB-21 | v-mAP@0.5: 63.1 -> 62.2 | Hurt |
| Use cyclic temporal coherency rather than non-attentive L2 alone (ablation) | Temporal coherence localization consistency · UCF101-24 | v-mAP@0.5: +0.3% | Helped |
| Use gradient smoothness attention rather than non-attentive L2 alone (ablation) | Localization consistency · UCF101-24 | v-mAP@0.5: +0.6% | Helped |
Closest papers by meaning
- Stable Mean Teacher for Semi-supervised Video Action Detection (AAAI 2025 · similarity 0.93)
- Semi-supervised Active Learning for Video Action Detection (AAAI 2024 · similarity 0.92)
- A flexible model for training action localization with varying levels of supervision (NeurIPS 2018 · similarity 0.89)
- TimeBalance: Temporally-Invariant and Temporally-Distinctive Video Representations for Semi-Supervised Action Recognition (CVPR 2023 · similarity 0.89)
- Similar Modality Enhancement and Action Consistency Learning for Weakly Supervised Temporal Action Localization (AAAI 2025 · similarity 0.87)
- Unsupervised Action Segmentation via Fast Learning of Semantically Consistent Actoms (AAAI 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.