15th Annual Conference of the International Speech Communication Association

September 14-18, 2014

Evaluation of Dictionary for Sparse Coding in Speech Processing

Yongjun He (1), Guanglu Sun (1), Guibin Zheng (2), Jiqing Han (2)

(1) Harbin Institute of Technology, China
(2) Harbin University of Science & Technology, China

As a promising technique, sparse coding has been widely used for the analysis, representation, compression, denoising and separation of speech. To represent a signal accurately and sparsely, a good dictionary which contains elemental signals is urgently desired and many methods have been proposed to obtain such a dictionary. However, there is a lack of reasonable evaluation methods to judge whether a dictionary is good enough. To solve this problem, we define a group of measures for the evaluation of a dictionary. These measures not only address the sparsity and reconstruction error in representation of a signal, but also consider the denoising and separating performances. Experiments show that the proposed measures can make reasonable evaluations.

Full Paper

Bibliographic reference.  He, Yongjun / Sun, Guanglu / Zheng, Guibin / Han, Jiqing (2014): "Evaluation of dictionary for sparse coding in speech processing", In INTERSPEECH-2014, 2361-2364.