A master thesis from Aalborg University
Multiple Static Segmentation of Videos using a Convolution of Mixtures of Gaussian processes
Term
4. term
Education
Publication year
2015
Submitted on
2015-06-07
Pages
72 pages
Abstract
In this project we work with multiple static segmentation of videos, which is the creation of multiple complementary segmentations of the pixels in a video. The work is based on a recently proposed method for multiple segmentation of image stacks and our own work on improving this method. We change the underlying probabilistic model from being non-temporal to being based on Gaussian processes. We experiment with different mean functions for the Gaussian process, and show how we can use a piecewise linear function as a mean function to obtain good segmentations. We find that the method is able find multiple interesting structures in videos, and also segmentations that we were unable to find with our previous method, it does however not give better results in all cases. We also do dimensionality reduction on the data using PCA, and show that we can obtain good results with a far shorter running time.
Documents
Colophon: This page is part of the AAU Student Projects portal, which is run by Aalborg University. Here, you can find and download publicly available bachelor's theses and master's projects from across the university dating from 2008 onwards. Student projects from before 2008 are available in printed form at Aalborg University Library.
If you have any questions about AAU Student Projects or the research registration, dissemination and analysis at Aalborg University, please feel free to contact the VBN team. You can also find more information in the AAU Student Projects FAQs.