CS
The Computer Science area of CIMAT is made up of a group of leading researchers at a national and international level in their work oriented towards research, teaching, training and technological development. The area carries out research on the following main topics:
Signal, image and computer vision processing
The Signal Processing and Computer Vision research group has four research areas: Signal and Image Processing and Analysis, Medical Image Processing, Computer Vision, and Multimedia Transmission.
In the area of Signal and Image Processing and Analysis, research is carried out in: A) Image restoration (noise reduction, deblurring, super-resolution,...) and image analysis (segmentation, classification of adversarial examples) using optimization and deep learning techniques, B) Detection of forest fires by determining fire and smoke, using image processing and machine learning techniques, C) Processing of satellite images (image registration, morphological reconstruction for monitoring urban cities), D) Image compression for efficient transmission and storage of information.
In the area of Medical Image Processing there are collaborations in: A) Adjustment of hydrogen diffusion models to magnetic resonance data to characterize neuronal connectivity and brain microstructure, B) quantitative estimation of myelin volume for disease detection and characterization of brain tissue microstructure, C) autofocus algorithms for microscopy applications, such as the analysis of mycobacterium tuberculosis.
In the area of Computer Vision, recent research covers topics related to the development of algorithms for the analysis of images of urban scenes: detection of objects/irregularities on the road, visual tracking, counting, trajectory prediction; Research is also carried out in 3D vision (scene reconstruction, SLAM...) with recent implicit or explicit representation techniques (Nerfs, Gaussian splatting).
In the area of Multimedia Transmission, the research focuses on the development of robust and secure schemes for real-time audio, image and video communication on the Internet. In particular, the objective is to optimize Internet resources (limited bandwidth) to maintain good quality of service, with the participation of the same users for the distribution of content. Under this scheme, communication systems or strategies have been developed based mainly on two technologies: A) Virtual Multicast (VM) and B) Peer-to-Peer (P2P). As fundamental research for the optimization of these technologies, novel schemes are being developed for real-time estimation of the delay between participants and available bandwidth. In the line of information protection, high-security and high-speed cryptographic schemes have been developed using chaotic dynamic systems.
The members of the group collaborate with research groups from various national and international institutions. Some of these institutions are: The Università degli Studi di Verona (Italy), the EPFL (Switzerland), the Institut National de Recherches en Informatique et en Automatique (INRIA, France), the University of California Riverside, the Institute of Optics of the Higher Council for Scientific Research (Spain), the Center for Mathematical Modeling (ModeMat) of the National Polytechnic University (Ecuador), the Institute of Neurobiology of UNAM, the Laboratory of Computational Learning and Image Research (CLIR) of the Faculty of Mathematics of the UADY, the Technological Institute of Mérida (ITM), the Center for Research in Geospatial Information Sciences (CentroGeo), the Urban Laboratory of the Model University (LU-UM), the Regional Health Research Center Public Health (CRISP) of the National Institute of Public Health (INSP).
Senior Researchers
Associate Researchers
Researchers for Mexico
Post-Doctorates
Organization of the Neuroimaging and Computational Vision Meeting for 25 years NeuroVisión together with the Institute of Neurobiology of UNAM. This meeting is the most important in the country, since it brings together computer vision methods and applications in medical images.
Pat-Chan, A. E., Hernandez-Lopez, F. J., & Moreno-Sabido, M. R. (2023). Object tracking in video using block matching method. Research in Computing Science, 152(8), pp. 21-33. ISSN: 1870-4069 Latindex. 1st place in the best poster award at the conference COMIA-2023.
Martínez-López, J. A., De-la-Torre-Gutierrez, H., & Hernandez-Lopez, F. J. (2023). Detection of urban antispaces using YOLO: case study Mexicali. Research in Computing Science, 152(9), pp. 21-33. ISSN: 1870-4069 Latindex. 3rd place in the best paper award at the conference COMIA-2023.
Organization of the Seminar on Applied Mathematics and Computing (SMAC) jointly with the Faculty of Mathematics of the UADY, which meets every 15 days inviting students and researchers to share their knowledge in order to promote their respective research work and create new collaborations.
The team led by CIMAT researchers (Alonso Ramírez-Manzanares, José L. Marroquín, Mariano Rivera et. al.) won the first place in computational modeling at the workshop White Matter Modelling Challenge at the International Symposium of Biomedical Imaging 2015, New-York (USA). The title of the work “Empirical Diffusion-and-Direction Distributions (ED^3) to Estimate White Matter Microstructure”.
PRONACES: 321075 “Integrated rural urban mobility and innovation in electromobility in Yucatan” of the 2021-2024 call National research and advocacy projects to transition to a socially and environmentally sustainable energy system. Project leader: Dr. Alfredo Arias Trinidad from the Engineering school of Modelo University (Merida, Mexico). Participation through collaborations. Achievements: 2 master thesis in progress, one from the postgraduate program in Statistical Computation at CIMAT-Monterrey, the other from the postgraduate program in Computer Science and Industrial Mathematics at CIMAT-Guanajuato.
Analysis of urban flow systems for the design of smart cities of the call to support scientific and technological research 2019 of Federal Technological Institutes and Centers. PNPC Postgraduate Modality. Technological Institute of Mérida. Jan-Dec/2019. Participation through collaborations. Achievements: Co-supervision of the thesis of two ITM master's students, “Vehicle traffic estimation in aerial video captures” and “Traffic sign recognition using artificial neural networks and augmented reality.”
Institutional participation in the subproject “Satellite Detection of Sargassum in Synthetic Aperture Radar Images”, within the project called “Sargassum Observation and Early Warning Systems (Pilot Project)”, supported by the National Strategic Science, Technology and Linkage Programs with the Social, Public and Private sectors of CONACyT, Project No. 321367, 2022.
Machine learning and data analysis
It covers learning based on variational and Bayesian methods, statistical learning, deep learning, natural language processing, multimodal learning, and applications of the above to computer vision, robotics, and engineering in general.
Senior Researchers
Associate Researchers
Researchers for Mexico
Post-Doctorates
Numerical methods, parallel computing and optimization
In the modeling, optimization, and parallel computing group, we focus on developing computational methods applicable to solving problems in engineering, physics, logistics and planning, among other areas. We are specialists in different methodologies involved in this type of problem, including system modeling, generally through differential equations, and optimization methods, including both classical and heuristic methods.
Nowadays, many problems requiring modeling and optimization use large amounts of data and computation. Our group deals with such issues by using parallel computing through the high-performance computing laboratories at CIMAT. We are experts in the development of parallel algorithms, including shared and distributed memory systems and GPUs.
One of the defining characteristics of our research group at CIMAT is our commitment to innovation. We don't usually rely on existing libraries for application development; we pioneer new techniques for simulation and optimization. This often leads to creating our own software and developing efficient and effective algorithms. As a result, we are also involved in developing algorithms and data structures that can be implemented efficiently and harness the power of recent hardware.
In the optimization area, research is carried out in both single-objective and multi-objective optimization, using stochastic optimization techniques such as evolutionary algorithms, memetic algorithms, estimation of distribution algorithms, among others. We also develop mathematical programming techniques and gradient methods. Both are applied to global and constrained design and optimization problems that arise in various fields, such as logistics, cryptography, or civil engineering. Through establishing teams with students interested in these topics, we have a tradition of participating in optimization contests to validate and show the advantages of the innovations made in this area. This has resulted in various international awards.
In the modeling area, numerical solution models of partial differential equations are studied and developed using Finite Elements, Finite Volume, Finite Differences, and, recently, Discrete Exterior Calculus (DEC). These techniques are used to solve problems in engineering applications. Studies are carried out on fuel cell design, thermal modeling in solids, modeling of fluids in physical systems, design of complex structures in civil engineering, development of new materials, biological material applications, and numerical modeling of groundwater and floods.
In parallel computing, we work on different parallelization technologies and paradigms, such as OpenMP, MPI, and CUDA, applying these tools to the parallelization of modeling and optimization algorithms. We are specialists in the efficient use of this type of hardware and offer consulting and training services.
The members of this group collaborate with various research groups from numerous national and international institutions. Some of the institutions with which we have collaborated are the following: CIMNE, University of La Laguna and University of Málaga, in Spain, Napier University, in Scotland, University of Leipzig, in Germany, University of Guanajuato, Autonomous University of Zacatecas , Universidad Michoacana San Nicolás de Hidalgo ITESM, CINVESTAV, CentroGeo, CICESE, ITAM, Instituto Tecnológico de Mérida, Universidad Autónoma de Yucatán, and CICY, in Mexico. Through these collaborations, it is possible to expand the set of problems and techniques available to the team and develop multidisciplinary teams to address the problems.
Senior Researchers
Associate Researchers
Researchers for Mexico
Post-Doctorates
Robotics and Intelligent Systems
The Robotics and Intelligent Systems (RSI by its acronym in Spanish) group at CIMAT has more than 15 years of experience conducting research in mobile robotics. Its theoretical approach, which transcends the practical, has made the CIMAT RSI group position itself as a reference within the robotics community in Mexico. The research works of its members are constantly published in the most important forums in the area, such as the IEEE Transactions on Robotics (IEEE TRO), International Journal of Robotics Research (IJRR), IEEE Robotics and automation Letters (IEEE RAL), Robotics and Autonomous Systems (RAS), Autonomous Robots (AURO), Automatica, ISA Transactions, IEEE Transactions on Control Systems Technology, Journal of the Franklin Institute, Control Engineering Practice, IEEE Transactions on Visualization and Computer Graphics, IEEE Transactions on Affective Computing, Virtual Reality, IEEE International Conference on Robotics and Automation (ICRA), International Workshop on the Algorithmic Foundations of Robotics (WAFR), International Conference on Intelligent Robots and Systems (IROS), among others.
The group is made up of twelve permanent members, all of them in the National System of Researchers (SNII by its acronym in Spanish) – one researcher with level 3, four with level 2, five with level 1, and two with candidate level. The research lines of the area focus globally on motion planning, control theory, perception, uncertainty modeling, and the integration of these different levels of estimation and decision through coherent and effective feedback schemes. More specifically, the most recent lines of research in the area have addressed problems through techniques such as optimal control, nonlinear control with sensory feedback, model predictive control, consensus-based distributed control, visual servo control, partially observable Markov decision processes, generative prediction models, Bayesian and combinatorial filtering, computational geometry, sampling-based planning methods, among others. The range of applications of the research covers ground and aerial robots, e.g. object search and reconstruction, exploration of unknown environments, multi-robot formations, pursuit/evasion, humanoid locomotion, autonomous cars and virtual reality.
The research group currently maintains international collaborations with Georgia Tech, U. de Sherbrooke, LAAS-CNRS, INRIA, U. of Oulu, Université de Technologie de Compiègne, U. de Zaragoza and national collaborations with Intel Labs Guadalajara, CINVESTAV Saltillo and Guadalajara, INAOE, CICESE, CIO Ags. and the U. of Guanajuato.
Senior Researchers
Associate Researchers
Researchers for Mexico
Post-Doctorates
1st place in LARC-VSS (Very Small Size Soccer) challenge, Mexican Robotics Tournament 2024.
2nd place in autonomous drone challenge, Mexican Robotics Tournament 2024.
2nd place in autonomous driving challenge: automodel car, Mexican Robotics Tournament 2024.
1st place in autonomous drone challenge, Mexican Robotics Tournament 2023.
Motion Strategies for Multiple Autonomous Robots in Unknown Crowded Environments, funded by Intel Corporation, as part of the collaboration with Intel Labs Guadalajara carried out between 2017 and 2019.
Vision-based Autonomous Landing in Populated Environments, a crucial aspect for urban deployment, funded by the Office of Naval Research, USA.
Postgraduate in
Computer
Science
Master of
Science with
orientation in
Robotics
Zacatecas Unit
Master in
Software
Engineering
Zacatecas Unit
Coordinator of the Computer Science Area
E-mail: dalmau@cimat.mx
Graduate Coordinator Computer Science Area
E-mail: alram@cimat.mx
