# PhD thesis defense: David Varas

David Varas, defends his PhD thesis entitled Region-based Particle Filter Leveraged with a Hierarchical Co-clusteringMonday November 14th, 11h30, Aula Telensenyament, B3 building, first floor |

Dissertation summary:

In this thesis, we exploit the hierarchical information associated with images to tackle two fundamental problems of computer vision: video object segmentation and video segmentation.

In the first part of the thesis, we present a video object segmentation approach that extends the well known particle filter algorithm to a region based image representation. Image partition is considered part of the particle filter measurement, which enriches the available information and leads to a reformulation of the particle filter theory. We define particles as unions of regions in the current image partition and their propagation is computed through a single optimization process. During this propagation, the prediction step is performed using a co-clustering between the previous image object partition and a partition of the current one, which allows us to tackle the evolution of non-rigid structures.

The second part of the thesis is devoted to the exploration of a co-clustering technique for video segmentation. This technique, given a collection of images and their associated hierarchies, clusters nodes from these hierarchies to obtain a coherent multiresolution representation of the image collection. We formalize the co-clustering as a Quadratic Semi-Assignment Problem and solve it with a linear programming relaxation approach that makes effective use of information from hierarchies. Initially, we address the problem of generating an optimal, coherent partition per image and, afterwards, we extend this method to a multiresolution framework. Finally, we particularize this framework to an iterative multiresolution video segmentation algorithm in sequences with small variations.

Finally, in the last part of the thesis we validate the presented techniques for object and video segmentation using the proposed algorithms as tools to tackle problems in a context for which they were not initially thought.