Patent · US Active

Systems and methods for gait recognition via disentangled representation learning

US11315363B2 · kind B2 · utility

0Cited by

7References

19Claims

0Family size

Assignees

Inventors

Xiaoming Liu · Schenectady, US
Jian Wan · Novi, US
Kwaku O. Prakah-Asante · Wixom, US
Mike Blommer · Ann Arbor, US
Ziyuan ZHANG · East Lansing, US
Luan Tran · Grand Rapids, US
Xi Yin · Sunnyvale, US
Yousef Atoum · East Lansing, US

Key dates

Filing date	Jan 22, 2021
Grant date	Apr 26, 2022
Priority date	—
Expiry date	Jan 22, 2041

Classification

Technology area (CPC G)Physics
CPC primaryG06V40/25
WIPO fieldComputer technology
WIPO sectorElectrical engineering

Abstract

Gait, the walking pattern of individuals, is one of the most important biometrics modalities. Most of the existing gait recognition methods take silhouettes or articulated body models as the gait features. These methods suffer from degraded recognition performance when handling confounding variables, such as clothing, carrying and view angle. To remedy this issue, a novel AutoEncoder framework is presented to explicitly disentangle pose and appearance features from RGB imagery and a long short-term memory integration of pose features over time produces the gait feature.

Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.