Research map / Time series / Foundation models for time series
HiMAE: Hierarchical Masked Autoencoders Discover Resolution-Specific Structure in Wearable Time Series
Simon A. Lee, Cyrus Tanade, Hao Zhou, Juhyeon Lee, Megha Thukral, Md Sazzad Hissain Khan, Keum San Chun, Baiying Lu, Migyeong Gwak, Mehrab Bin Morshed, Viswam Nathan, Md Mahbubur Rahman and 3 more
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 hierarchical design and skip connections (ablation) | HiMAE · masked reconstruction scaling | masked MSE | Improved |
| use hierarchical convolutional U-Net instead of transformer LSM | masked autoencoder · parameter scaling, reconstruction | masked MSE | Improved |
| use HiMAE against transformer baselines | wearable SSL · random imputation, interpolation, extrapolation | MSE | Improved |
| use HiMAE embeddings against larger foundation models | wearable SSL · 12 binary classification tasks, linear probe | AUROC: not first on some tasks, about 1-2% behind winner | Mixed |
| probe different hierarchy levels of HiMAE (ablation) | HiMAE · downstream PPG tasks | AUROC by layer | Mixed |
| choose patch length 5 and kernel size 5 (ablation) | HiMAE · pretraining hyperparameters | reconstruction | Improved |
| use compact CNN instead of 110M transformers for on-watch inference | HiMAE · PVC detection on watch CPU | model size and compute: 4.8 MB vs 423 MB and 441 MB | Improved |
Closest papers by meaning
- Scaling Wearable Foundation Models (ICLR 2025 · similarity 0.83)
- RelCon: Relative Contrastive Learning for a Motion Foundation Model for Wearable Data (ICLR 2025 · similarity 0.82)
- VCR: Learning Valid Contextual Representation for Incomplete Wearable Signals (NeurIPS 2026 · similarity 0.81)
- MoPFormer: Motion-Primitive Transformer for Wearable-Sensor Activity Recognition (NeurIPS 2025 · similarity 0.81)
- Beyond Sensor Data: Foundation Models of Behavioral Data from Wearables Improve Health Predictions (ICML 2025 · similarity 0.81)
- Scalable Hierarchical Deep Neural Network for Time Series Analysis in Wearable Sensor-based Human Activity Recognition (AAAI 2025 · similarity 0.81)
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.
