Radar-APLANC: Unsupervised Radar-based Heartbeat Sensing via Augmented Pseudo-Label and Noise Contrast
Ying Wang, Zhaodong Sun, Xu Cheng, Zuxian He, Xiaobai Li · 2025
Read Radar-APLANC: Unsupervised Radar-based Heartbeat Sensing via Augmented Pseudo-Label and Noise Contrast by Ying Wang, Zhaodong Sun, Xu Cheng, Zuxian He, Xiaobai Li (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2511-08071.txt.
Description
Ying Wang and colleagues introduce Radar-APLANC, an unsupervised method for sensing heartbeats with radar and no contact. Frequency modulated radar can pick up tiny chest movements, but noise degrades older signal processing methods. Learning based methods resist noise better, yet they usually need costly labeled recordings for training.
The framework builds positive samples from the heartbeat range and negative samples from noise ranges in the radar matrix. A noise contrastive triplet loss trains on pseudo labels from a traditional method, so no true physiological labels are needed. Adaptive label selection cleans those pseudo labels, and results rival supervised methods on two data sets.
Catalog record
| CREATOR | Ying Wang, Zhaodong Sun, Xu Cheng, Zuxian He, Xiaobai Li |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 30 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2511-08071