CCG/ZGDV Institute participated in the 6th Advanced Computing Meeting, held on 3–4 November 2026 at the University of Minho, in Braga, Portugal.
As part of the “BSC & CCG AI Factory – Computer Vision” workshop, held on 3 November, researchers Diogo Lopes and Edel Reyes delivered the presentation “From High-Performance Computing to Applied Research in Computer Vision”, showcasing four research applications that illustrate the potential of High-Performance Computing (HPC) to accelerate scientific research and technological innovation. The initiative also featured Susana Caetano (FCT/AI²) and provided an opportunity to share knowledge, experiences and practical applications of Computer Vision, fostering interaction between researchers, participants and CCG/ZGDV specialists.
The four case studies addressed challenges across three application domains:
Healthcare – Multispectral Imaging and Digital Pathology
The development of AI models for occult blood detection and Bristol stool classification using multispectral imaging, alongside foundation models for digital pathology. HPC supports self-supervised learning, processing gigapixel whole-slide images, and developing reusable visual representations, with the aim of reducing dependence on extensive expert annotations.
Industry – Synthetic X-ray Imaging for Weld Inspection
The generation of synthetic radiographic images of welded joints through physics-based simulation. HPC enables large-scale simulations across different materials, pipe geometries, and acquisition conditions, contributing to the creation of diverse, traceable datasets for industrial inspection.
Infrastructure Inspection – Intelligent 3D Reconstruction of Bridges
The integration of LiDAR, IMU and GNSS/RTK data for trajectory estimation, sensor fusion and 3D reconstruction of bridges. Advanced computing enables the processing of large point clouds and computationally intensive optimisation algorithms, supporting accurate spatial reconstruction even in environments with degraded GNSS signals.
Beyond these applications, the presentation addressed how HPC resources can be effectively integrated into research workflows, from selecting computational resources and parallel execution to efficient, reproducible scientific computing. The central message is that HPC is not simply about more computing power. It enables researchers to explore larger datasets, more complex models and experimental approaches that would otherwise be impractical with conventional computing resources.
Through its research in Computer Vision and Artificial Intelligence, CCG/ZGDV Institute continues to strengthen the connection between advanced computing, scientific innovation and real-world applications, contributing to the development of intelligent solutions for healthcare, industry and infrastructure inspection.
The 6th Advanced Computing Meeting, organised by FCT through FCCN, in partnership with the University of Minho and CNCA, brought together researchers, technology specialists, companies and institutional representatives to exchange experiences and discuss the future of advanced computing in Portugal and Europe.
