Monday, September 23, 2013

Bayesian Robust Matrix Factorization for Image and Video Processing - implementation -

Naiyan Wang just sent me the following:

Dear Igor,

We just published a new paper titled "Bayesian Robust Matrix Factorization for Image and Video Processing" in ICCV13'. In this paper, we give a full Bayesian formulation to robust matrix factorization (a.k.a Robust PCA) problem, and extend the model to handle contiguous outliers which are often encountered in computer vision applications. The paper, supplemental material and codes are all available at:

Thanks Naiyan !

The full page presenting this impressive implementation is here and the paper is: Bayesian Robust Matrix Factorization for Image and Video Processing by Naiyan Wang and Dit-Yan Yeung (Supplemental Material here)

Matrix factorization is a fundamental problem that is often encountered in many computer vision and machine learning tasks. In recent years, enhancing the robustness of matrix factorization methods has attracted much attention in the research community. To benefit from the strengths of full Bayesian treatment over point estimation, we propose here a full Bayesian approach to robust matrix factoriza-tion. For the generative process, the model parameters have conjugate priors and the likelihood (or noise model) takes the form of a Laplace mixture. For Bayesian inference, we devise an effiient sampling algorithm by exploiting a hierarchical view of the Laplace distribution. Besides the basic model, we also propose an extension which assumes that the outliers exhibit spatial or temporal proximity as encoun-tered in many computer vision applications. The proposed methods give competitive experimental results when compared with several state-of-the-art methods on some benchmark image and video processing tasks
The implementation is here. Maybe Cable and I should use in our adventures and use it on the Lana del Rey dataset. Anyway, the solver will be added shortly to the Advanced Matrix Factorization Jungle page.

No comments: