Towards multi-scale fuzzy sparse discriminant analysis using local third-order tensor model of face images

Published: 2016, Last Modified: 13 Nov 2024Neurocomputing 2016EveryoneRevisionsBibTeXCC BY-SA 4.0
Abstract: Highlights•Exploit a set of multi-scale characteristics using Ridgelet transform.•Solve Ridgelet transformed coefficients with respect to each image block.•Synthesize a new coefficient vector for each training image.•Evaluate the sparse similarity grade between each training sample and class.•Redesign a new discriminant criterion to perform robust classification.
Loading