Utilizing a hybrid model to recognize fast and precise hand inputs in a virtual environment
US10956724B1 · kind B1 · utility
Assignee
Inventor
Key dates
| Filing date | Sep 10, 2019 |
| Grant date | Mar 23, 2021 |
| Priority date | — |
| Expiry date | Sep 10, 2039 |
Classification
- Technology area (CPC G)Physics
- CPC primaryG06N3/08
- WIPO fieldComputer technology
- WIPO sectorElectrical engineering
Abstract
The present disclosure relates to systems, non-transitory computer-readable media, and methods for improving hand input detection in a virtual environment based on small and/or rapid hand inputs. For instance, a hybrid hand input detection system can utilize a hybrid hand input model that incorporates a machine-learning hand pose model with an optical flow network to recognize fine degrees of hand movement corresponding to input commands within a virtual environment. In this manner, the hybrid hand input detection system can utilize a hybrid hand input model to recognize small and rapid hand inputs in virtual reality or augmented reality without requiring a physical input device.
Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.