Online Blind Separation of Dependent Sources Using Nonnegative Matrix Factorization Based on KL Divergence
نویسندگان
چکیده
This paper proposes a novel online algorithm for nonnegative matrix factorization (NMF) based on the generalized Kullback-Leibler (KL) divergence criterion, aimed to overcome the high computation problem of large-scale data brought about by conventional batch NMF algorithms. It features stable updating the factors alternately for each new-coming observation, and provides an efficient solution for the blind separation of statistically dependent sources (i.e., the sources are mutually correlated). Our theoretic analysis is validated by simulation examples. Streszczenie. Przedstawiono nowy algorytm do faktoryzacji nieujemnej macierzy bazujący na kryterium Kullback-Leibler, pozwalający usprawnić problem obliczeń dużej ilości danych. Algorytm sukcesywnie zmienia współczynniki i pozwala na ślepą separację statystycznie zależnych źródeł. (On-line ślepa separacja zależnych źródeł przy użyciu faktoryzacji nieujemnej matrycy bazująca na kryterium Kullback-Leibler)
منابع مشابه
Pattern Expression Nonnegative Matrix Factorization: Algorithm and Applications to Blind Source Separation
Independent component analysis (ICA) is a widely applicable and effective approach in blind source separation (BSS), with limitations that sources are statistically independent. However, more common situation is blind source separation for nonnegative linear model (NNLM) where the observations are nonnegative linear combinations of nonnegative sources, and the sources may be statistically depen...
متن کاملA Modified Digital Image Watermarking Scheme Based on Nonnegative Matrix Factorization
This paper presents a modified digital image watermarking method based on nonnegative matrix factorization. Firstly, host image is factorized to the product of three nonnegative matrices. Then, the centric matrix is transferred to discrete cosine transform domain. Watermark is embedded in low frequency band of this matrix and next, the reverse of the transform is computed. Finally, watermarked ...
متن کاملA Modified Digital Image Watermarking Scheme Based on Nonnegative Matrix Factorization
This paper presents a modified digital image watermarking method based on nonnegative matrix factorization. Firstly, host image is factorized to the product of three nonnegative matrices. Then, the centric matrix is transferred to discrete cosine transform domain. Watermark is embedded in low frequency band of this matrix and next, the reverse of the transform is computed. Finally, watermarked ...
متن کاملExperimental Evaluation of Superresolution-Based Nonnegative Matrix Factorization for Binaural Recording
In this paper, we propose a new stereo signal separation scheme based on multi-divergence supervised nonnegative matrix factorization (SNMF). In previous studies, a hybrid method, which concatenates superresolutionbased SNMF after directional clustering, has been proposed for multichannel signal separation. However, the optimal divergence in SNMF temporally fluctuates because the separation and...
متن کاملFast Local Algorithms for Large Scale Nonnegative Matrix and Tensor Factorizations
Nonnegative matrix factorization (NMF) and its extensions such as Nonnegative Tensor Factorization (NTF) have become prominent techniques for blind sources separation (BSS), analysis of image databases, data mining and other information retrieval and clustering applications. In this paper we propose a family of efficient algorithms for NMF/NTF, as well as sparse nonnegative coding and represent...
متن کامل