Research map / Video / Action recognition
Unsupervised Learning and Segmentation of Complex Activities From Video
Fadime Sener, Angela Yao
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 the GMM temporal ordering model (ablation) | the appearance model on Inria Instructional Videos without background · Inria Instructional Videos without background | Mof: 0.322 -> 0.692 | Improved |
| use the fully unsupervised iterated method | Breakfast Actions · Breakfast Actions | Mof: 33.3 -> 34.6 | Improved |
| use the fully unsupervised iterated method | Breakfast Actions · Breakfast Actions | Jaccard: 47.3 -> 47.1 | No clear effect |
| increase rho_0 to 5 to encourage a stricter canonical ordering | the proposed GMM model · Inria Instructional Videos, change tire | F1: 0.41 vs. 0.42 | No clear effect |
| use a few Gaussian mixture components for the visual appearance model (ablation) | the proposed model · Inria Instructional Videos without background | Mof | Improved |
| vary the embedding dimensionality (ablation) | the proposed model · Inria Instructional Videos without background | Mof | No clear effect |
| account for background frames in the model (ablation) | the standard model on Inria Instructional Videos · Inria Instructional Videos | Mof | Improved |
| use the GMM ordering model on activities with repeated sub-activities | the proposed model · Inria Instructional Videos, perform CPR | F1 | Hurt |
| increase the assumed number of sub-activities K (ablation) | the proposed model · Inria Instructional Videos without background | Mof | No clear effect |
Closest papers by meaning
- Unsupervised Learning of Action Classes With Continuous Temporal Embedding (CVPR 2019 · similarity 0.89)
- Unsupervised Action Segmentation by Joint Representation Learning and Online Clustering (CVPR 2022 · similarity 0.89)
- Weakly-Supervised Online Action Segmentation in Multi-View Instructional Videos (CVPR 2022 · similarity 0.88)
- Unsupervised Action Segmentation via Fast Learning of Semantically Consistent Actoms (AAAI 2024 · similarity 0.86)
- Hierarchical Vector Quantization for Unsupervised Action Segmentation (AAAI 2025 · similarity 0.86)
- Temporally Consistent Unbalanced Optimal Transport for Unsupervised Action Segmentation (CVPR 2024 · 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.
