The group is composed of several researchers from the departments of Computer Science and Statistics and Research. Operational with a long history of working together, along with the incorporation of other people who arrived later to the Computer Science department and several collaborators, also linked by previous joint research, from the Universitat Jaume I of Castellón. The most general common nexus is computer vision and image analysis, both 2D and recently 3D, with a special focus on medical imaging and the one generated by biological processes. The common goal is to provide experience, curriculum and applicable solutions to medical or industrial problems related to image analysis, shape analysis, reconstruction and modelling of anatomical structures and retrieval of information from image databases. Due to the complexity of the software that must be developed, a formal vision which deals with the modelling of the software and its interaction with the user is necessary. In particular, the research activity carried out to date, and which is intended to be given even greater cohesion, is organised along the following lines:
- Segmentation and co-registration of anatomical structures, in particular from radiology images, magnetic resonance images, positron emission tomography (PET) images or other modalities, if required. The statistical analysis of the shapes obtained for their comparison, indexing or modelling requires the use of morphometric techniques that connect this line to the next.
- Morphometry, understood as the statistical analysis of shapes, both 2D and 3D, to determine their temporal variations or between groups of cases and to obtain representative prototypes of shape classes.
- Computational Physiology, understood as the multi-scale modelling of biological and medical processes by means of ICT tools to better understand pathophysiology and improve the diagnosis and treatment of diseases. Larger scale modelling (organs) connects this line to the previous one, as long as shape analysis is applied to organs such as the liver or the heart; smaller scale modelling connects it to the next line,
- Stochastic spatio-temporal models for the analysis of dynamic processes from image sequences. In particular, statistical methodologies based on univariate and bivariate germ-grain processes that have been used so far to model processes in cell biology is applied by analysing images generated by confocal microscopy.
- Image and shape retrieval based on the visual content of large image or shape databases, in general, not manually labelled with a descriptive text, with special focus on human morphometry and medical image databases. This supports the organisation and semantic description of the case studies used in previous lines.
- Software production methods and modelling of software interaction with the user. This transversal line analyses and arranges the software produced (in fact, the main applicable outcome of our research) so that it is correct, reusable, extensible and easily manageable (in the case of final products) by content-competent but non-computer-specialised users, in particular doctors or health personnel. The applicability of this research is ocused on the biomedical area, and has its ultimate goal in clinical application. However, there are interesting derivations in fields such as basic research in areas like cell biology, or material science and other more direct practical utilities (design of communication and sensor networks, clothing design, shopping recommendation systems that use the visual aspect of objects, etc.).
The ISP research group, http://isp.uv.es, has a long tradition in statistical analysis of data coming from imaging systems. These measurements depend on the properties of the scenes and the physics of the imaging process, and their relevance depends on the (natural or artificial) observer that will analyze the data. Our distinct approach to signal, image and vision processing combines machine learning theory with the understanding of the underlying physics and biological vision. Applications mainly focus on optical remote sensing and computational visual neuroscience. Empirical statistical inference, also known as machine learning, is a field of computer science interested in making predictions, and models from observations and sensory data. The information processing tools in machine learning are critical to understand the function of natural neural networks involved in biological vision, as well as to make inferences in complex dynamic network systems, such as the Earth biosphere, atmosphere, and ecosystems.
Problems in Visual Neuroscience and in Remote Sensing based geosciences require similar mathematical tools. For example, both scientific fields face model inversion and model understanding problems. In both cases, one has a complex forward model that is difficult to invert (to extract information from) either because it is not analytically invertible (undetermined) or because the measurements (or responses) are noisy in nature. In Remote Sensing, the forward model is the imaging process given certain state conditions in the surface and atmosphere. In Visual Neuroscience, the forward model includes what is known in the neural pathway from the retina to the different regions of the visual cortex. Inversion of such models is key to make quantitative and meaningful inferences about the underlying system that generated the observed data. Beyond such quantitative assessment, a qualitative interpretation of the proposed models is mandatory as well. Qualitative understanding is more challenging than prediction, and causal inference from empirical data is the common playground both in geoscience and neuroscience. Simultaneous observations and recordings from a phenomenon lead to multidimensional signals that may display strong statistical correlation between the components. However, correlation is not enough to establish cause-effect relationships. This is key when analyzing activation and inhibition in the communication between different brain regions, and it is also of paramount relevance when studying the causes, effects and confounders of essential climate variables for detection and attribution in climate science. Finally, another parallelism is the analysis of big visual data: hyperspectral imagery acquired by current and upcoming satellite sensors pose a big-data information processing problem in similar ways to that in the visual brain. Adaptation, pattern recognition, inference and decision making in the brain may be quite inspiring for remote sensing image analysis.
The group is therefore organized into a theoretical research branch (A) and a more applied research branch (B). The theoretical machine learning core tackles model inversion, interpretation, causal inference from empirical data and inclusion of physical constraints and prior knowledge in big visual data. The applied research lines are devoted to apply and adapt the theoretrical developments for remote sensing, geociences and visual neuroscience. For the sake of simplicity, we have grouped together these activities along five conceptual research lines: machine learning, visual neuroscience, image processing, remote sensing and big data processing.
The LISITT group was set up in 1989 with the aim of filling the existing gap in Spain in the area of telematics applications in the field of traffic and transport. Its initial activities were focused on the execution of international research and development projects within the European ESPRIT and DRIVE programmes of the 2nd Framework Programme of the European Union.
Since its origins, LISITT has specialised in the study and development of Intelligent Transport Systems (ITS), covering technological, organisational and strategic aspects. LISITT has been carrying out projects for more than 20 years for national traffic and transport administrations, including the Directorate General of Traffic, the Ministry of Public Works and its regional counterparts in the Basque and Catalan Governments. LISITT is currently a multidisciplinary group (Physics, Civil Engineering, Computer Engineering, Telecommunications Engineering, Mathematics, Geography) that brings together more than 60 professionals, all of them university graduates, including civil servants, contracted teachers and its own research staff, and has established itself as a reference group in consultancy on telematics applied to transport, in the development of ITS systems, and strategic consultancy on management issues and the development of traffic systems.
The work carried out since its origins has consolidated LISITT as a Spanish reference group in consultancy on telematics applied to transport, in the development of ITS systems, and strategic consultancy on management, development and maintenance of traffic systems for administrations, as reflected by the fact that LISITT has been participating for more than 10 years as expert advisors representing the Directorate-General for Traffic in different national and international standardisation committees and in European working groups on ITS systems, including the World Committee for Standardisation in ITS systems ISO/TC204, the European Committee for Standardisation of ITS systems CEN/TC278 and the Spanish Committee for Telematics applied to transport and road traffic AEN/CTN 159. The role played by LISITT in the creation, assistance and monitoring of the Euro-regional SERTI project (1995 - 2006), the Euro-regional ARTS project (1997 - 2006) and the European EasyWay project (2007-2013) should also be highlighted.
Apart from these consultancy activities in the standardisation groups in the field of ITS systems, LISITT's most important projects are grouped around the following topics:
- Consultancy to traffic administrations on coordination and organisation of international traffic control and management projects.
- Technical assistance to public administrations in traffic management and information systems.
- Study, development and maintenance of traffic information systems for public traffic administrations.
- Coordination and execution of R&D&I projects, both from the European Union and national calls for proposals.
- Analysis, design, construction and development of information systems for private companies.
- Computer security, data protection and privacy.
The main purpose of IDAL is the study and application of intelligent methods of data analysis for pattern recognition, with applications that struggle with prediction, classification or trend determination.
Its members apply classic statistical methods and automatic learning techniques to large databases: statistical hypothesis testing, linear models, feature selection and extraction, neural networks, clustering algorithms, decision trees, support vector machines, probabilistic graphical models, manifold visualization, fuzzy logic, reinforcement learning, etc.
The ultimate goal of the application of these methods is to generate mathematical models which enable the optimization of processes and resources, as well as to reach the optimal decision making stage. A clear example of this is the area of health, where IDAL has developed clinical decision support application based on data analysis. These applications make it possible to improve the patient’s quality of life (establishing optimal clinical guidelines) while reducing healthcare costs.
Complementing this knowledge, the group has extensive experience in signal processing (spectral analysis, digital filter, adaptive process, etc.) due to their work of over 10 years in biosignal processing (mainly ECG and EEG). With all this background, IDAL is able to analyse a wide range of data and signals. This fact is backed up by the large number of both private and public contracts it has developed in different areas of knowledge. Furthermore, most of the practical work carried out has been displayed in important scientific publications with high impact parameters and in a large number of communications to international congresses within the area of data analysis.
Among the developed applications, (outside the health area already mentioned) are the following, i.a: web recommendations, models for optimal incentive management to gain customer loyalty, measurement-based shoe recommendations, and other data analysis consultancy works. In addition to its practical work IDAL, it develops new data analysis algorithms improving the performance of the existing ones. This research work is also reflected in a wide dissemination in the form of different publications in journals of impact and in congresses of data analysis relevant to the scientific community.
The research group in Public Opinión and Elections aims at analyzing, studying and finding solutions to all issues and questions related to electoral processes and/or the measurement and monitoring of public opinion, applying the most advanced quantitative techniques.
The most relevant research fields of the group include (but are not limited to) the following: the generation of electoral predictions, inference of individual voting behavior, analysis of polls and surveys, the search of new methodological approaches to improve (reducing costs) the quality of sampling methods, semantic analysis of opinions and monitoring of the internet sentiment, the study of the consequences of non-response and of the biases introduced during the whole inference process, the solution to the gaps in the databases, the integration and pooling of local and global information to obtain multilevel responses, and the development of statistcal theory and methodology.
The approach used in the research group in Public Opinión and Elections is open, not being limited by any particular methodological tendency, and makes extensive use of whatever sources of information. Thus, we use classical and Bayesian techniques, we apply from simple linear regression models to complex approaches based on neural networks, wavelets or auto-binomial models, we use the spatial and/or temporal component of the data explicitly, we perform simulation via Markov chain Monte Carlo or directly by Monte Carlo methods, and we introduce in our models survey data, reported election results, news reports, internet messages and/or official statistics.
The members of the group are open to working with other research groups, companies and institutions and encourage interested parties to contact us in order to explore possible avenues of collaboration.
The research group LSyM focuses its activity on simulation systems development, employing the latest Virtual Reality techniques. LSyM has always worked looking for close collaboration with the company and obtaining important results in the field of civil works. The group is a part of the Institute on Robotics and Information and Communications Technologies (IRTIC) of the Universitat de València.
Lines of research:
- Integration of the real-time immersive simulators: design of all simulator’s elements, including both hardware and software (dynamic models of objects and 3-D scenarios).
- Development of e-learning platforms based on 3-D simulation: simulation technologies based on WebGL and Unity-3D in order to implement virtual 3-D environments, executable from the browser on different computing platforms. Use of Moodle and other e-learning standards
- Advanced computing in graphics processing units (GPUs): Development of performance league calculation programmes based on CUDA, OpenCL and shaders, that run on GPU network architectures.
- Real-time physical modelling: Development of simulation and of models of collaborative behaviour between avatars.
Fields of application:
- Industrial: Virtual and augmented reality systems in several industrial areas (transport, railway sector, construction, maritime sector, etc.).
- Education: Web-based simulation of learning environments, e-learning platforms for training and evaluation courses.
Services to companies and other entities:
Technical assistance and consulting on:
- Development of real-time virtual environments to train operators of industrial machinery, cranes, civil engineering machinery and vehicles.
- Counselling on the integration of low-medium-high cost simulators and on the choice of the appropriate hardware for the app.
- Design and implementation of training systems based on the use of simulators in different areas (transport, heavy machinery, air traffic controllers, etc.) and focused on learning risk prevention techniques.
- Development of e-learning platforms
