people
members of the lab or group
LISA - Laboratories of Image, Signal processing and Acoustic Image Research Unit
Av. F.D. Roosevelt 50, CP 165/56 - 1050 Bruxelles
layout: about permalink: /about/ title: about nav: true nav_order: 1
profile: align: right image: prof_pic.jpg image_circular: false # crops the image to make it circular more_info: > <p>LISA - Laboratories of Image, Signal processing and Acoustic Image Research Unit</p> <p>Av. F.D. Roosevelt 50, CP 165/56 - 1050 Bruxelles - Belgium</p> —
About
My research focuses on real-time 3D vision, computational imaging, and immersive media. I develop algorithms and software that enable photorealistic view synthesis, accurate depth estimation, and efficient scene representations for applications ranging from virtual and augmented reality to robotics and telepresence.
I obtained a master’s degree in Computational Intelligence and Robotics Engineering from Université libre de Bruxelles and completed a joint Ph.D. at Université libre de Bruxelles and Vrije Universiteit Brussel. My research combines computer vision, computer graphics, optimization, and high-performance computing to address the challenges of interactive 3D visual computing.
A central aspect of my work has been the development of reference software for the Moving Picture Experts Group (MPEG). Three of my software implementations have been adopted as MPEG reference software and are used by both academic and industrial participants in the standardization process. My research has also contributed to the European HoviTron project, where advanced view synthesis is integrated into CREAL’s light-field display technology to enable immersive teleoperation of robotic systems.
My principal research contribution is the Reference View Synthesizer (RVS), a real-time view synthesis framework capable of producing photorealistic novel views while maintaining a remarkably small computational footprint. RVS demonstrates that high visual quality can be achieved without the heavy computational requirements typically associated with neural rendering methods, making it suitable for real-time applications and standardization activities.
Complementing RVS, I developed the Reference Depth Estimation (RDE) software, a high-precision depth estimation framework designed to generate accurate depth maps for high-quality rendering. Since depth estimation remains one of the limiting factors in view synthesis, this work addresses both reconstruction quality and computational efficiency.
Another research direction concerns plenoptic imaging. My work includes methods for extracting sub-aperture images, estimating intrinsic camera parameters without calibration patterns, and exploiting plenoptic 2.0 cameras for dense micro-baseline acquisition. These imaging systems naturally complement view synthesis by providing densely sampled viewpoints that improve reconstruction quality and enable novel forms of image-based rendering.
Beyond view synthesis and computational imaging, I have worked on real-time point cloud rendering using splatting techniques, Gaussian-process optimization, deep-learning approaches for view synthesis, robotic acquisition platforms capable of sub-millimetre precision, and high-performance software engineering. My experience spans Linux and Windows environments, GPU programming, Docker, PostgreSQL, and modern web technologies.
My research philosophy is to combine theoretical advances with robust software implementations that are practical enough to be adopted by both the research community and industry. I believe that impactful research should not only advance the state of the art, but also provide reproducible tools that enable others to build upon it.
As my research continues to evolve, I remain interested in efficient algorithms for immersive visual computing and in bridging the gap between fundamental research, standardization, and industrial deployment.
LISA - Laboratories of Image, Signal processing and Acoustic Image Research Unit
Av. F.D. Roosevelt 50, CP 165/56 - 1050 Bruxelles
layout: about permalink: /about/ title: about nav: true nav_order: 1
profile: align: right image: prof_pic.jpg image_circular: false # crops the image to make it circular more_info: > <p>LISA - Laboratories of Image, Signal processing and Acoustic Image Research Unit</p> <p>Av. F.D. Roosevelt 50, CP 165/56 - 1050 Bruxelles - Belgium</p> —
About
My research focuses on real-time 3D vision, computational imaging, and immersive media. I develop algorithms and software that enable photorealistic view synthesis, accurate depth estimation, and efficient scene representations for applications ranging from virtual and augmented reality to robotics and telepresence.
I obtained a master’s degree in Computational Intelligence and Robotics Engineering from Université libre de Bruxelles and completed a joint Ph.D. at Université libre de Bruxelles and Vrije Universiteit Brussel. My research combines computer vision, computer graphics, optimization, and high-performance computing to address the challenges of interactive 3D visual computing.
A central aspect of my work has been the development of reference software for the Moving Picture Experts Group (MPEG). Three of my software implementations have been adopted as MPEG reference software and are used by both academic and industrial participants in the standardization process. My research has also contributed to the European HoviTron project, where advanced view synthesis is integrated into CREAL’s light-field display technology to enable immersive teleoperation of robotic systems.
My principal research contribution is the Reference View Synthesizer (RVS), a real-time view synthesis framework capable of producing photorealistic novel views while maintaining a remarkably small computational footprint. RVS demonstrates that high visual quality can be achieved without the heavy computational requirements typically associated with neural rendering methods, making it suitable for real-time applications and standardization activities.
Complementing RVS, I developed the Reference Depth Estimation (RDE) software, a high-precision depth estimation framework designed to generate accurate depth maps for high-quality rendering. Since depth estimation remains one of the limiting factors in view synthesis, this work addresses both reconstruction quality and computational efficiency.
Another research direction concerns plenoptic imaging. My work includes methods for extracting sub-aperture images, estimating intrinsic camera parameters without calibration patterns, and exploiting plenoptic 2.0 cameras for dense micro-baseline acquisition. These imaging systems naturally complement view synthesis by providing densely sampled viewpoints that improve reconstruction quality and enable novel forms of image-based rendering.
Beyond view synthesis and computational imaging, I have worked on real-time point cloud rendering using splatting techniques, Gaussian-process optimization, deep-learning approaches for view synthesis, robotic acquisition platforms capable of sub-millimetre precision, and high-performance software engineering. My experience spans Linux and Windows environments, GPU programming, Docker, PostgreSQL, and modern web technologies.
My research philosophy is to combine theoretical advances with robust software implementations that are practical enough to be adopted by both the research community and industry. I believe that impactful research should not only advance the state of the art, but also provide reproducible tools that enable others to build upon it.
As my research continues to evolve, I remain interested in efficient algorithms for immersive visual computing and in bridging the gap between fundamental research, standardization, and industrial deployment.