A late fusion multi-task learning for respiratory waveform and rate estimation from photoplethysmography
Minh Nhut Ho, Kien Trong Nguyen · 2026
Read A late fusion multi-task learning for respiratory waveform and rate estimation from photoplethysmography by Minh Nhut Ho, Kien Trong Nguyen (2026) 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/plos-pone-0353203.txt.
Description
Ho and Nguyen built a learning system that reads breathing from a fingertip light sensor in a 2026 study. The sensor, already common for pulse checks, carries breathing clues in how its wave rises and falls. Earlier systems estimated only breathing rate or only the wave shape, never both at once.
Separate branches read each kind of clue, then merge to produce both outputs together. Stepwise training, where the system adapts from one reference signal type to another, beat all other strategies. Mixing reference types carelessly hurt both outputs. Short sensor windows thus can track breathing fully, if training respects which reference the labels came from.
Catalog record
| CREATOR | Minh Nhut Ho, Kien Trong Nguyen |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 53828 |
| LICENSE | CC_BY |
| RIGHTS | CC BY |
| SOURCE | PLOS |