A customized AI-based machine learning model for evaluating participants’ activities under a workshop setting
Xiaoyi Shao, Hua Li, Qian Ma, Kai Jin · 2026
Read A customized AI-based machine learning model for evaluating participants’ activities under a workshop setting by Xiaoyi Shao, Hua Li, Qian Ma, Kai Jin (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/PMC13518891.txt.
Description
Xiaoyi Shao, Hua Li, Qian Ma, and Kai Jin build a 2026 machine learning system that recognizes what participants are doing during hands-on workshops. Action recognition struggles here because activities overlap, class counts are lopsided, and meaning depends on context. Their framework fine-tunes a 3D convolutional network in stages, then adds an unsupervised review.
Training used 208 clips from internal and external datasets, plus public UCF101 material for a side analysis. The tailored 57-class model separated most actions well within its home context, yet similar actions blended and cross-context transfer suffered from merged categories and drift. Shao and colleagues suggest adversarial alignment and pose guidance as next fixes.
Catalog record
| CREATOR | Xiaoyi Shao, Hua Li, Qian Ma, Kai Jin |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 54243 |
| LICENSE | https://creativecommons.org/licenses/by/4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |