Research map / Label-efficient and robust learning / Semi-supervised learning
LabelForest: Non-Parametric Semi-Supervised Learning for Activity Recognition
Yuchao Ma, Hassan Ghasemzadeh
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 silhouette-based sample filtering after greedy spanning forest (balanced, 1 seed per class) (ablation) | LabelForest · HART selective labeling | precision: +15.4% precision, labeling rate 16.6% of GSF | Improved |
| add silhouette-based sample filtering after greedy spanning forest (unbalanced, 8 seeds in one class) (ablation) | LabelForest · HART selective labeling | precision: +17.5% precision, labeling rate 18.8% of GSF | Improved |
| add sample filtering (smaller but cleaner labeled set) for downstream training (ablation) | LabelForest · HART activity recognition | accuracy | Improved |
| increase seed samples per label from 1 to 6 | LabelForest · labeling accuracy, three datasets | labeling accuracy: +7.2%, +8.3%, +9.8% | Improved |
| increase seed samples per label | LabelForest · HART activity recognition | accuracy: 74.6% -> 91.1% | Improved |
| use unbalanced rather than balanced seeds (L8 vs L2) | LabelForest · selective labeling, three datasets | labeling accuracy: -3.6% on average | Hurt |
Closest papers by meaning
- Distribution-Based Semi-Supervised Learning for Activity Recognition (AAAI 2019 · similarity 0.89)
- Multi-Label Activity Recognition Using Activity-Specific Features and Activity Correlations (CVPR 2021 · similarity 0.87)
- Joint Prediction of Activity Labels and Starting Times in Untrimmed Videos (ICCV 2017 · similarity 0.87)
- Weakly-Guided Self-Supervised Pretraining for Temporal Activity Detection (AAAI 2023 · similarity 0.86)
- Learning to Recognize Human Activities Using Soft Labels (TPAMI 2016 · similarity 0.86)
- Domain-Informed Label Fusion Surpasses LLMs in Free-Living Activity Classification (Student Abstract) (AAAI 2025 · 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.
