Dendron: Enhancing Human Activity Recognition with On-Device TinyML Learning
Hazem Hesham Yousef Shalby, Manuel Roveri · 2025
Read Dendron: Enhancing Human Activity Recognition with On-Device TinyML Learning by Hazem Hesham Yousef Shalby, Manuel Roveri (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-2503-01353.txt.
Description
Hazem Hesham Yousef Shalby and Manuel Roveri present Dendron, a tiny on-device learning method that names human activities from wearable sensors. Their paper from twenty twenty-five sits in human activity recognition. Recent work runs the naming on the device itself, under the TinyML banner. Learning new tasks on the device, though, was missing, most when labeled data are scarce.
Dendron fills that gap for small embedded boards. It picks up fresh activity classes from few supervised samples, right where the data are sensed. Tests use two public data sets and an off-the-shelf board named in the paper. Scores and costs both look strong in those runs. The work suits health and sport wearables that must adapt to their owners. Privacy gains too, since data need never leave the device.
Catalog record
| CREATOR | Hazem Hesham Yousef Shalby, Manuel Roveri |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 34 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-01353