Cross view gait recognition using joint-direct linear discriminant analysis

摘要: This paper proposes a view-invariant gait recognition framework that employs unique view invariant model profits from the dimensionality reduction provided by Direct Linear Discriminant Analysis (DLDA). The framework, which energy images (GEIs), creates single joint accurately classifies GEIs captured at different angles. Moreover, proposed also helps to reduce under-sampling problem (USP) usually appears when number of training samples is much smaller than dimension feature space. Evaluation experiments compare framework's computational complexity and accuracy against those other methods. Results show improvements in both accuracy.