Computational Learning and Imaging Research
Welcome to the Laboratory for Computational Learning and Imaging Research (CLIR) of the Universidad Autónoma de Yucatán. The main objective of CLIR is to develop innovative computational solutions in the fields of machine learning, imaging, and computer vision.
Research at CLIR in machine learning covers a wide range of topics going from applications of classical machine learning algorithms for solving challenging problems in fields such as computer vision, robotics and pattern analysis of bio-metric data, to the theoretical study and design of new learning models. A considerable amount of effort in CLIR is aimed to developing variational learning models.
The arrival of the Internet of Things also presents new opportunities and challenges to test learning algorithms over large data-sets. Although it is true that vast amount of computational resources are required, it is also known that a careful engineered design of the algorithm's implementation is critical to achieve high performance. Research on efficient parallel and distributed computing by means of CPU/GPU clusters for collaborative computing is also performed at CLIR.
In general, we aim to contribute to the academic community by offering a number of machine learning and related courses, supervising dissertations/theses and performing state of the art research with the objective to tackle specific demands of the general public.
Imaging and Computer vision
In CLIR we conduct extensive research in computer vision techniques aimed to solve challenging problems in robotics, optical metrology and the processing of digital images.
Computer vision aims to provide machines a way to interpreting their surroundings by means of the information collected from a set of cameras which observe the neighboring objects and structures present in the scene. Our group has a collection of works on state of the art image processing methods such as image denoising, inpainting, segmentation, registration, phase recovery using structured light triangulation models, pattern analysis and learning methods applied to medical images.
A very interesting sub-field of computer vision is the one of developing algorithms for embedded systems. This means investigating the holistic optimization of the hardware and the software in order to find methods to obtain affordable, high-performance, and energy-efficient solutions often with real-time computing constraints. In these lines, CLIR has been supported by funding provided by the Program for the Professional Teaching Development (PRODEP) to carry out multidisciplinary research to explore autonomous and cooperative control algorithms for UAV teams capable of flying in dynamic and uncertain environments.