I presented a talk on June the 21st, 2018 :
How to use Gaussian mixture models on patches for solving image inverse problems
during the MixStatSeq workshop.
The slides of the talk can be found here.
Je présente le modèle HDMI au séminaire doctorant du LAMFA à Amiens mardi 24 octobre.
Abstract: In this talk, we first present a statistical framework for patch-based image denoising. This framework requires the inference of statistical models in high dimensional spaces which leads to several challenges due to the curse of dimensionality. To tackle this, we propose a model with intrinsic dimensionality reduction which yields state-of-the-art results in image denoising.
I’m presenting a talk about my latest resarch High-Dimensional Mixture Models for Unsupervised Image Denoising (HDMI) in the minisymposium Boosting and Learning in Mathematical Imaging Algorithms at the SIAM annual meeting 2017 in Pittsburgh, Pennsylvania.
You can find the slides of the presentation here.