Title | ||
---|---|---|
A Transform Domain Implementation of Sparse Representation Method for Robust Face Recognition |
Abstract | ||
---|---|---|
Recently, a new discriminative sparse representation method for robust face recognition via
$$\textit{l}_2$$
regularization (NDSRFR) was reported. In this contribution, a transform domain (TDNDSRFR) is presented. The discrete cosine transform implementation of the TDNDSRFR is given and shown to maintain the recognition accuracy of the NDSRFR while yielding considerable reduction in the computational complexity and storage requirements. Also, a technique that selects the balance parameter is introduced. Extensive simulations were performed on six face databases, namely, ORL, YALE, FERET, FEI, Cropped AR, and Georgia Tech, and sample results are given which confirm the enhanced properties of the TDNDSRFR. |
Year | DOI | Venue |
---|---|---|
2019 | 10.1007/s00034-019-01099-w | Circuits, Systems, and Signal Processing |
Keywords | Field | DocType |
Discrete cosine transform, Sparse representation, Face recognition | Georgia tech,Facial recognition system,Mathematical optimization,Pattern recognition,FERET,Discrete cosine transform,Sparse approximation,Regularization (mathematics),Artificial intelligence,Discriminative model,Mathematics,Computational complexity theory | Journal |
Volume | Issue | ISSN |
38 | 9 | 0278-081X |
Citations | PageRank | References |
0 | 0.34 | 0 |
Authors | ||
2 |
Name | Order | Citations | PageRank |
---|---|---|---|
Taif Alobaidi | 1 | 1 | 3.06 |
Wasfy B. Mikhael | 2 | 76 | 76.27 |