Visual Scene Reconstruction Using a Bayesian Learning Framework

In this paper, we focus on constructing new flexible and powerful parametric framework for visual data modeling and reconstruction. In particular, we propose a Bayesian density estimation method based upon mixtures of scaled Dirichlet distributions. The consideration of Bayesian learning is interesting in several respects. It allows simultaneous parameters estimation and model selection, it permits also taking uncertainty into account by introducing prior information about the parameters and it allows overcoming learning problems related to over- or under-fitting. In this work, three key issues related to the Bayesian mixture learning are addressed which are the choice of prior distributions, the estimation of the parameters, and the selection of the number of components. Moreover, a principled Metropolis-within-Gibbs sampler algorithm for scaled Dirichlet mixtures is developed. Finally, the proposed Bayesian framework is tested on a challenging real-life application namely visual scene reconstruction.

[1]  Nizar Bouguila,et al.  MML-Based Approach for Finite Dirichlet Mixture Estimation and Selection , 2005, MLDM.

[2]  Jean-Michel Marin,et al.  Bayesian Modelling and Inference on Mixtures of Distributions , 2005 .

[3]  Nizar Bouguila,et al.  Bayesian Learning of Generalized Gaussian Mixture Models on Biomedical Images , 2010, ANNPR.

[4]  Peter Congdon,et al.  Applied Bayesian Modelling , 2003 .

[5]  Geoffrey J. McLachlan,et al.  Finite Mixture Models , 2019, Annual Review of Statistics and Its Application.

[6]  Sami Bourouis,et al.  Color image segmentation with bounded generalized Gaussian mixture model and feature selection , 2018, 2018 4th International Conference on Advanced Technologies for Signal and Image Processing (ATSIP).

[7]  Nizar Bouguila,et al.  Finite Generalized Gaussian Mixture Modeling and Applications to Image and Video Foreground Segmentation , 2007, Fourth Canadian Conference on Computer and Robot Vision (CRV '07).

[8]  Sami Bourouis,et al.  Bayesian learning of finite generalized inverted Dirichlet mixtures: Application to object classification and forgery detection , 2014, Expert Syst. Appl..

[9]  Michael Goesele,et al.  Scene Reconstruction and Visualization From Community Photo Collections , 2010, Proceedings of the IEEE.

[10]  Jean-Yves Guillemaut,et al.  General Dynamic Scene Reconstruction from Multiple View Video , 2015, 2015 IEEE International Conference on Computer Vision (ICCV).

[11]  Nizar Bouguila,et al.  Novel approaches for synthesizing video textures , 2012, Expert Syst. Appl..

[12]  Nizar Bouguila,et al.  Variational learning of a Dirichlet process of generalized Dirichlet distributions for simultaneous clustering and feature selection , 2013, Pattern Recognit..

[13]  Andrew W. Fitzgibbon,et al.  Image-Based Rendering Using Image-Based Priors , 2005, International Journal of Computer Vision.

[14]  Nizar Bouguila,et al.  Bayesian hybrid generative discriminative learning based on finite Liouville mixture models , 2011, Pattern Recognit..

[15]  Nizar Bouguila,et al.  A countably infinite mixture model for clustering and feature selection , 2011, Knowledge and Information Systems.

[16]  Sami Bourouis,et al.  A hierarchical Dirichlet process mixture of generalized Dirichlet distributions for feature selection , 2015, Comput. Electr. Eng..

[17]  Sami Bourouis,et al.  Image and video denoising by combining unsupervised bounded generalized gaussian mixture modeling and spatial information , 2018, Multimedia Tools and Applications.

[18]  Irfan A. Essa,et al.  Machine Learning for Video-Based Rendering , 2000, NIPS.

[19]  Fatma Najar,et al.  A Comparison Between Different Gaussian-Based Mixture Models , 2017, 2017 IEEE/ACS 14th International Conference on Computer Systems and Applications (AICCSA).

[20]  Baoxin Li,et al.  Probabilistic Image-Based Rendering with Gaussian Mixture Model , 2006, 18th International Conference on Pattern Recognition (ICPR'06).

[21]  Nizar Bouguila,et al.  On online high-dimensional spherical data clustering and feature selection , 2013, Eng. Appl. Artif. Intell..

[22]  Nizar Bouguila,et al.  Unsupervised learning of finite mixtures using scaled dirichlet distribution and its application to software modules categorization , 2017, 2017 IEEE International Conference on Industrial Technology (ICIT).

[23]  Nizar Bouguila,et al.  A Model-Based Approach for Discrete Data Clustering and Feature Weighting Using MAP and Stochastic Complexity , 2009, IEEE Transactions on Knowledge and Data Engineering.