IBPDL-SVA

Linux & osx	Quality	Code Review

This repository implements the IBP-DL SVA algorithm from [dang2018,elvira2018].

Prerequirements

Our instructions have been tested on Linux and Mac only. On Mac, you may need to install a compiler, e.g., gcc (as part of the XCode command line tools).

Install from sources

1. Clone this repository

git clone https://github.com/c-elvira/IBPDL-SVA.git
cd IBPDL-SVA

And run

./install.sh

to complete the installation.

2. Test

The installation can be tested using the following simple command

python setup.py test

Usage examples

Work in Progress :)

Reproducible research

The folders 'exp_eusipco' and 'exp_cap' contain the code to reproduce the experiments in [dang2018,elvira2018].

This work is associated to the following papers

@Inproceedings{elvira2018,
    title     = {Small variance asymptotics and bayesian nonparametrics for dictionary learning},
    author    = {Elvira, Clément and Dang, Hong-Phuong and Chainais, Pierre},
    booktitle = {Proc. European Signal Processing Conf. (EUSIPCO)},
    address   = {Rome, Italy},
    year      = {2018},
    month     = {Sept.},
}

@Inproceedings{dang2018,
    title     = {Vers une méthode d'optimisation non paramétrique pour l'apprentissage de dictionnaire en utilisant Small-Variance Asymptotics pour modèle probabiliste},
    author    = {Dang, Hong-Phuong and Elvira, Clément and Chainais, Pierre},
    booktitle = {(CAP)},
    address   = {Rouen, France},
    year      = {2018},
    month     = {Juin},
}