Translator Disclaimer
1 August 2011 Enhanced discriminative locality alignment and its kernel extension
Author Affiliations +
Although discriminative locality alignment (DLA), which is based on the idea of part optimization and whole alignment, has better performance than classical methods in feature extraction, DLA is too overly sensitive to the values of the parameters and falls short of exploiting the full supervision information. We propose a novel supervised feature extraction method, named enhanced discriminative locality alignment (EDLA), for robust feature extraction. EDLA is not sensitive on the choice of the parameters, and both the local structure and class label information are taken into consideration in EDLA algorithm. Moreover, a kernel version of EDLA, named kernel EDLA, is developed through applying the kernel trick to EDLA to increase its performance on nonlinear feature extraction. Experiments on the face databases demonstrate the effectiveness of our methods.
©(2011) Society of Photo-Optical Instrumentation Engineers (SPIE)
Zhonghua Liu and Zhong Jin "Enhanced discriminative locality alignment and its kernel extension," Optical Engineering 50(8), 087002 (1 August 2011).
Published: 1 August 2011


Face recognition with L1-norm subspaces
Proceedings of SPIE (May 04 2016)
Content-based indexing for medical image databases
Proceedings of SPIE (October 01 1998)
Multiple experts for robust face authentication
Proceedings of SPIE (April 01 1998)

Back to Top