University of Cambridge > Talks.cam > Machine Learning @ CUED > Non-Parametric Conditional Random Fields in Computer Vision and Image Processing

Non-Parametric Conditional Random Fields in Computer Vision and Image Processing

Add to your list(s) Download to your calendar using vCal

If you have a question about this talk, please contact Novi Quadrianto.

In this talk, I will discuss ongoing work on non-parametric random fields at the computer vision group of Microsoft Research Cambridge. The models I will present can be considered as conditional random fields, the clique potentials of which depend linearly on the model parameters, but non-linearly on the input features. The main focus of the talk will be on how such non-linear maps from features to clique potentials can be learned efficiently and jointly with the model parameters, under a common objective function. I will present such models for both continuous and discrete output variables, yielding an empirically successful framework for many structured prediction tasks arising in computer vision and image processing. The utility of the approach will be exemplified by means of several applications where our approach achieves state-of-the-art results, including denoising, deblurring/deconvolution and image inpainting.

This talk is part of the Machine Learning @ CUED series.

Tell a friend about this talk:

This talk is included in these lists:

Note that ex-directory lists are not shown.

 

© 2006-2017 Talks.cam, University of Cambridge. Contact Us | Help and Documentation | Privacy and Publicity