Extended common molecular and discriminative atom dictionary based sparse representation for face recognition
| Autor: | Zheng-ping Hu, Fan Bai, Shuhuan Zhao, Meng Wang, Zhe Sun |
|---|---|
| Rok vydání: | 2016 |
| Předmět: |
K-SVD
Computer science business.industry Representation (systemics) 020206 networking & telecommunications Pattern recognition 02 engineering and technology Sparse approximation Class (biology) Facial recognition system Discriminative model Simple (abstract algebra) Face (geometry) Signal Processing 0202 electrical engineering electronic engineering information engineering Media Technology 020201 artificial intelligence & image processing Computer Vision and Pattern Recognition Artificial intelligence Electrical and Electronic Engineering business |
| Zdroj: | Journal of Visual Communication and Image Representation. 40:42-50 |
| ISSN: | 1047-3203 |
| Popis: | An extended common molecular and discriminative atom dictionary based sparse representation method is proposed.We propose a very simple way to produce a multiple view representation dictionary.Face recognition is performed by combination of common information and discriminative information.We demonstrated that adding the common information is effective. The employed dictionary plays an important role in sparse representation classification, however how to build the relationship between dictionary atoms and class labels is still an important open question. Many existing sparse representation classification dictionary models exploit only the discriminative information either in the representation coefficients or in the representation residual, which limits their performance. To address this issue, we introduce a novel dictionary building method which is constructed by two parts: the common molecular dictionary and the discriminative atom dictionary. More specifically, the discriminative atom dictionary builds its relationship to class labels and the extended molecular dictionary can reduce the representation residual for all the classes. Therefore, the new dictionary not only has correspondence to the class labels, but also has the perfect representation ability. Besides, the maximum probability representation is used for the final classification. In conclusion, the sparse coefficient of our method is sparser than the sparse representation-based classification (SRC), and our method can achieve better performance. Experiments on the AR, Extended Yale B and CMU PIE face datasets verify that our algorithm outperforms many recently proposed methods. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full Text from ScienceDirect