Research map / Label-efficient and robust learning / Recognition and long-tail
Instilling Social to Physical: Co-Regularized Heterogeneous Transfer Learning
Read the paperCheck 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 |
|---|---|---|---|
| Add social media knowledge via naive feature concatenation | Sensor-only features · HAR | accuracy | Helped |
| Use word-embedding label alignment instead of topic-model (LDA) alignment (ablation) | CoHTL · HAR | accuracy | Helped |
| Use CoHTL matrix factorization instead of DAMA manifold assumption | heterogeneous alignment · HAR | accuracy | Helped |
| Increase beta co-regularization too much (ablation) | CoHTL · HAR 10-fold CV | average accuracy: best at β = 0.1 | Hurt |
Closest papers by meaning
- Limitations in Employing Natural Language Supervision for Sensor-Based Human Activity Recognition - And Ways to Overcome Them (AAAI 2025 · similarity 0.83)
- Generalizable Sensor-Based Activity Recognition via Categorical Concept Invariant Learning (AAAI 2025 · similarity 0.83)
- SensorLM: Learning the Language of Wearable Sensors (NeurIPS 2025 · similarity 0.82)
- Domain-Informed Label Fusion Surpasses LLMs in Free-Living Activity Classification (Student Abstract) (AAAI 2025 · similarity 0.81)
- HARDVS: Revisiting Human Activity Recognition with Dynamic Vision Sensors (AAAI 2024 · similarity 0.81)
- TraXion: Rethinking Pre-training Frameworks for Mobility and Beyond (NeurIPS 2026 · similarity 0.80)
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.