Can I read Face Recognition Using Tensors and Gradient Descriptors on EtoBox?

Face Recognition Using Tensors and Gradient Descriptors by Oprea, Cristina C. ;Dogariu, Laura M. is a scholarly article available to read on EtoBox.

What is Face Recognition Using Tensors and Gradient Descriptors about?

Face recognition algorithms are key issues in many applications, still challenging in terms of complexity and reliability. In this paper we propose a different approach for a solution based on tensors and multilinear singular value decomposition (MLSVD). We define a new tensorization of face image descriptors built into gradient space and rely on such feature vectors to identify a subject from the database. The tensor is decomposed using MLSVD model to form a linear set of equations with a Kronecker product solution. We show that our method lowers the computation complexity while keeping high accuracy in the recognition process.

Author

Oprea, Cristina C. ;Dogariu, Laura M.

Publisher

IEEE

Published

2020

Language

EN