Visual and Data-Intensive Computing

The Visual and Data-intensive Computing (ViDiC) Sector is active in the research, development, and application of innovative and scalable solutions for acquiring, creating, processing, distributing, exploring, and analyzing complex and/or massive datasets derived from simulations or measurements of phenomena, objects, environments, or real-world processes.

Specifically, our work focuses on advancing methods in the following areas.

• Scalability of methods and technologies for ingesting, processing, managing, and analyzing large datasets or data streams. This broad area focuses on studying, developing, and applying state-of-the-art techniques in HPC, distributed computing, automation, and their integration to enhance the efficient, reproducible, secure, and trustworthy utilization of available data.
  Main keywords: big data, automation and orchestration, workflows, reproducibility, high-performance computing, and distributed computing.
• Mathematical modeling, numerical simulation, data aggregation and analysis, and, where relevant, machine learning and artificial intelligence, to understand and solve complex problems for predictive and interventional purposes.
  Main keywords: mathematical modeling, numerical simulation, machine learning, physics-informed networks, and demand & response.
• Modeling, integration, and management of heterogeneous data and interoperability between processes in different application domains through semantics-preserving open standards and formalisms. The aim is to support the collaborative acquisition, sharing, and analysis of complex data, improving their understanding and reuse through a specific focus on their meaning and generation context, on traceability at every stage of the process, and on sophisticated usage consent management.
  Main keywords: Findability - Accessibility - Interoperability - Reusability (FAIR), standardization, semantic modeling, data integration and management, traceability and provenance.
• Visual and geometric computing to process, analyze, and synthesize visual and spatial data. The goal is to enable computers to transform simulated or measured data into visual or geometric information and to develop scalable and intuitive methods that allow users to explore and interact with massive and/or complex visual and geometric content.
  Main keywords: computer graphics, computational geometry, scientific and information visualization, computer vision, geometric and visual machine learning, 3D/XR/VR/AR, and digital fabrication. See also [VIC].

Although our work is generally cross-cutting to many application domains, we are particularly active in the fields of biomedicine, precision medicine, and clinical informatics, including digital and computational pathology, digital biobanking, genomics, and in silico modeling of chemical-physical and biological phenomena at the molecular and cellular levels for medical and pharmacological applications. We are also highly active in urban informatics and cultural heritage sciences, particularly in the smart reconstruction of structured environments, shapes, and materials, and the interactive exploration of complex and annotated models. We also strongly contribute in the fields of imaging and computational geophysics, as well as in the energy and environment fields, with a focus on meteorology and climatology, and the forecasting and control of electricity production and consumption in the context of the energy transition. Our activities are primarily conducted within the framework of collaborations with external partners, including international organizations, industries, universities, research centers, research hospitals, museums, and other cultural organizations, with a strong emphasis on public-interest outcomes and the dissemination of knowledge, including through the development of standards and best practices.