Locality-sensitive dictionary learning for sparse representation based classification

• Published in: Pattern recognition 2013-05, Vol.46 (5), p.1277-1287
• Main Authors: Wei, Chia-Po, Chao, Yu-Wei, Yeh, Yi-Ren, Wang, Yu-Chiang Frank
• Format: Article
• Language: English
• Subjects: Algorithms; Applied sciences; Classification; Coding, codes; Data locality; Dictionaries; Dictionary learning; Exact sciences and technology; Exact solutions; Face recognition; Image processing; Information, signal and communications theory; Learning; Pattern recognition; Preserves; Representations; Signal and communications theory; Signal processing; Signal representation. Spectral analysis; Signal, noise; Sparse representation; Telecommunications and information theory
• ISSN: 0031-3203; 1873-5142
• DOI: 10.1016/j.patcog.2012.11.014

Motivated by image reconstruction, sparse representation based classification (SRC) has been shown to be an effective method for applications like face recognition. In this paper, we propose a locality-sensitive dictionary learning algorithm for SRC, in which the designed dictionary is able to preserve local data structure, resulting in improved image classification. During the dictionary update and sparse coding stages in the proposed algorithm, we provide closed-form solutions and enforce the data locality constraint throughout the learning process. In contrast to previous dictionary learning approaches utilizing sparse representation techniques, which did not (or only partially) take data locality into consideration, our algorithm is able to produce a more representative dictionary and thus achieves better performance. We conduct experiments on databases designed for face and handwritten digit recognition. For such reconstruction-based classification problems, we will confirm that our proposed method results in better or comparable performance as state-of-the-art SRC methods do, while less training time for dictionary learning can be achieved. ► A locality-sensitive dictionary learning algorithm is proposed for SRC. ► Our method preserves data locality and thus results in improved classification. ► Closed-form solutions are derived for the proposed method for fast convergence. ► Experiments on face and digit recognition problems support the use of our method. ► We achieve better/comparable performance as state-of-the-art SRC methods do.