Video-based hand gesture recognition via SPD manifold spatial representation and optical flow motion features
Zhonghai Bai, Václav Snášel, Seyedali Mirjalili, Bay Vo, Jeng-Shyang Pan, Lingping Kong, Crina Grosan · 2026
Read Video-based hand gesture recognition via SPD manifold spatial representation and optical flow motion features by Zhonghai Bai, Václav Snášel, Seyedali Mirjalili, Bay Vo, Jeng-Shyang Pan, Lingping Kong, Crina Grosan (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-0348122.txt.
Description
Zhonghai Bai and a large team combined two views of a moving hand to recognise gestures from video more reliably. One view captures the hand's shape through statistical descriptors treated with special geometry that respects their curved mathematical home. The other tracks motion, summarising how brightness flows across a grid from frame to frame. Either view alone misses part of the story.
Fused together, the combined description beat both single views on two public benchmark sets, scoring above ninety-nine percent on one and above ninety-seven on the other. The representation stays compact and its parts remain interpretable, which helps developers debug mistakes. The authors aim the framework at touchless controls, sign language tools, and other systems where a camera has to read hands in real time.
Catalog record
| CREATOR | Zhonghai Bai, Václav Snášel, Seyedali Mirjalili, Bay Vo, Jeng-Shyang Pan, Lingping Kong, Crina Grosan |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 64615 |
| LICENSE | CC_BY |
| RIGHTS | CC BY |
| SOURCE | PLOS |