Our lab, in collaboration with CYENS Centre of Excellence, features state-of-the-art facilities and equipment designed to support groundbreaking research in deep learning, computer vision, augmented and virtual reality, computer graphics, and more. Some Key Infrastructure Highlights:

  1. Virtual Production Studio
    • Equipped with a 30 sqm LED wall delivering more than 100% coverage of the REC2020 color space, ultra-high resolution, and low latency, ideal for immersive media applications.
  2. Motion Capture Systems
    • Conventional Motion Capture: A 24-camera PhaseSpace Inc. Impulse X2E system with active markers, capturing at 960fps and full 13MP resolution. This system utilizes modulated LEDs and linear scanner arrays to accurately track 3D motion.
    • Portable Wireless Mocap Systems: Multiple XSENS and Rokoko systems, including face and hand capture capabilities, suitable for both lab-based and field use.
    • Volumetric Capture: We have installed the Evercoast Volumetric Capture system with 24 cameras.
  3. Film Production and High-Performance Computing
    • Film production equipment paired with state-of-the-art computing systems, including:
      • Supermicro servers with 80TB high-speed storage.
      • Workstations featuring AMD Threadripper CPUs and NVidia A6000 GPUs for real-time, hyper-realistic graphics rendering on platforms like Unreal Engine.
  4. Extended Reality (XR) Devices
    • A diverse range of head-mounted devices, including Meta Quest, HTC Vive Pro, and Microsoft HoloLens, enabling immersive experiences.
  5. AI and Deep Learning Infrastructure
    • Access to an advanced AI cluster comprising 8 computational nodes, each equipped with 8 NVIDIA Ampere GPUs, offering exceptional performance for AI-based applications.
  6. Additional Specialized Equipment
    • 3D printers and specialized scanners, such as insect-specific scanners, to support diverse research needs.

Our comprehensive infrastructure empowers researchers and creators to push the boundaries of technology and innovation in a collaborative environment.

Cutting-Edge Lab Infrastructure

In our lab, we are investigating novel and effective methods in character animation, at the intersection of computer vision and graphics. More specifically, we are interested in methods and applications in the wider area of character animation, including pose and motion reconstruction, motion analysis, emotion recognition, style tranfer, motion retargeting, motion synthesis, and motion interaction. More recently, we have developed innovative frameworks using deep and generative AI models; we are also interested in working with reinforcement learning, and physics-based animations.     

Character Animation

Motion capture (MoCap) is a cutting-edge technology used to acquire precise three-dimensional (3D) positioning and orientation data of moving objects. Its ability to capture and accurately portray complex movements has made it invaluable across various fields, including entertainment, media, military, sports, rehabilitation, and medical sciences.

At our lab, we have continuously researched, designed, and developed advanced algorithms for pose reconstruction using computer vision and sparse sensor data. Additionally, we focus on motion capture techniques to achieve full-body motion reconstruction, enabling detailed analysis and replication of articulated motion with exceptional accuracy.

Motion Capture and Pose Reconstruction

Our lab is actively engaged in developing advanced crowd simulations to analyze behavioral patterns, populate VR/AR environments with realistic virtual agents, model emergency evacuation scenarios, and reconstruct 3D motion trajectories from 2D data for diverse research and practical applications.

Crowd Simulations

Our lab is dedicated to advancing the design and development of immersive VR, AR, and XR environments, focusing on a wide range of innovative applications. These include the creation of digital twins for simulating real-world systems, the development of virtual applications for diverse industries, the preservation and recreation of digital heritage for cultural and educational purposes, and the modeling and populating of virtual cities and spaces. Through these efforts, we aim to push the boundaries of technology, blending creativity with cutting-edge research to address complex challenges and open new possibilities for virtual experiences.

VR/AR/XR Environments

Over the last few decades, a number of methods have been proposed to record, e-document, preserve, protect and disseminate mostly tangible cultural heritage. Apart thought from the tangible artifacts, cultural heritage also encompasses a range of important intangible assets that includes collective knowledge of communities, skills, practices, expressions, art, fashion and representations that do not have a tangible form. Intangible Cultural Heritage (ICH) is a mainspring of humanity’s cultural diversity and its maintenance is a guarantee for continuing creativity. In this direction, we are investigating methods and developing applications that contribute in the hollisting acquisition, documentation, reconstruction, analysis, synthesis, and visualization of both tangible and intangible creations, and including but not limited in folk dancing and musical instruments. 

Digital Cultural Heritage

Geometric Algebra (GA) provides a convenient mathematical notation for representing orientations and rotations of objects in three dimensions. The conformal model of GA (CGA) is a mathematical framework that offers a compact and geometrically intuitive formulation of algorithms, and an easy and immediate computation of rotors. Rotors are simpler to manipulate than Euler angles, and they are more numerically stable and efficient than rotation matrices, avoiding the problem of gimbal lock. 

Geometric algebra is particularly well suited to allow cross-disciplinary solutions in software engineering as it provides an intuitive and insightful common denominator across mathematical disciplines used in a variety of applications. Understanding GA enables us to relate distinct, seemingly incompatible paths by providing a common geometric and mathematical base. CGA gives us also the ability to describe algorithms in a geometrically intuitive and compact manner, making it suitable for applications in engineering, computer vision, graphics, and robotics.

Geometric Algebra

We offer several courses related to computer graphics, machine learning, character animation, computer games, computer vision, and motion capturing.  

Courses at the University of Cyprus