Menu Content/Inhalt
Computer Vision & Learning Group (CVL-UGR) PDF Print E-mail

The Computer Vision & Learning Group (CVL-UGR) is comprised of researchers of the groups:

Visual Image Processing (VIP) Group from the University of Granada (UGR).

Uncertainty Treatment in Artificial Intelligence (UTAI) Group from the University of Granada (UGR).

Intelligent Systems Group (ISG) from the University of the Basque Country (EHU).

The research interests of the CVL-UGR group lie in the following areas:

Image restoration and superresolution
Image restoration tries to eliminate the blur, noise and compression artifacts present in image and video sequences captured by digital cameras due to poor capture conditions and high compression ratios used for transmission or storage. Superresolution techniques aim to obtain high quality high resolution images from a set of low resolution images or low resolution video sequences.

Shape and human-action recognition
We approach the Human Action Recognition task from several image low-level information. In particular we focus on local histogram of optical flow, border, gradient, etc. Using machine learning multilayer architectures (RBM, DBN) we learn middle and high level features with the highest discriminative power.

Probabilistic graphical models
Probabilistic graphical models have been the main methodological tool to deal with uncertainty in different fields such as artificial intelligence, data mining, computer vision, bioinformatics, etc. The research groups UTAI y ISG have developed different approaches to use probabilistic graphical models, specifically Bayesian networks, in supervised classification, clustering and feature subset selection.

Information retrieval
Information retrieval is the interdisciplinary science of searching for documents and for information within documents which satisfy the information needs of users expressed by means of queries. Probabilistic models, specially models for structured XML document retrieval, are of our interest.

Evolutionary computation
Evolutionary computation compiles a set of algorithms inspired in the evolution of species in nature and devoted to learning and optimization problems. Estimation of Distribution Algorithms developed by IS Group, which make use of explicit probability models and has been probed to be usefull in machine learning, bioinformatics, computer vision, etc.

CVL-UGR Members