Projects
Deep Learning Methods for Global Illumination Transport in Computer Graphics KU Leuven
The global illumination light transport problem is a longstanding problem in computer graphics. Given the description of a 3D virtual scene, a photo-realistic image is generated. The state-of-the-art algorithms are all based Monte Carlo integration techniques. Recent advances in image synthesis have included machine learning principles such that the computational effort per pixel can be reduced. These techniques are often aimed at stochastic ...
Increasing the convergence speed of psysically-based rendering algorithms in computer graphics using frequency-based combined adaptive sampling and adaptive filtering KU Leuven
Physically-based rendering is one of the major research topics in the field of computer graphics. It deals with the generation of photo-realistic images depicting a virtual world. Although research in this field has a history dating back more than 25 years, generating photo-realistic images efficiently is still a problem of major significance. For each pixel in the final image a light transport equation needs to be evaluated, of which the ...
Tractable Approximations for Achieving Higher Model Efficiency in Computer Vision KU Leuven
The 2010s have seen the first large-scale successes of computer vision “in the wild”, paving the way for industrial applications. Thanks to the formidable increase of processing power in consumer electronics, convolutional neural networks have led the way in this revolution. With enough supervision, these models have proven able to surpass human accuracy on many vision tasks. However, rather than focusing exclusively on accuracy, it is ...
A modular consistent, disciminative framework for structured output learning in computer vision. KU Leuven
State of the art computer vision systems are fundamentally reliant on statistical learning to optimize performance on a specific application. Currently, statistical frameworks in computer vision are typically based on classification and regression, probabilistic graphical models, or discriminative structured prediction frameworks such as the Structured Output Support Vector Machine (SOSVM). Although some of the best performing ...
Using Generated Data to Train Computer Vision Models. Hasselt University
A Modular, Consistent, Discriminative Framework for Structured Output Learning in Computer Vision. KU Leuven
State of the art computer vision systems are fundamentally reliant on statistical learning to optimize performance on a specific application. Currently, statistical frameworks in computer vision are typically based on classification and regression, probabilistic graphical models, or discriminative structured prediction frameworks such as the Structured Output Support Vector Machine (SOSVM). Although some of the best performing computer vision ...
Prediction of gait neuromechanics following orthopedic interventions in children with cerebral palsy using computer simulations based on personalized models KU Leuven
Children with cerebral palsy (CP) suffer from a brain lesion that leads to impaired motor control, spasticity and muscle weakness. All these factors undermine the subjects' gait performance and, with time, will pose limitation to their mobility, independence and self-care. Orthopedic interventions aim at improving the walking performance. However, functional outcomes are not always as expected and, often, follow up surgery is needed to ...
A combined GP-GPU/FPGA desktop system for accelerating image processing applications (GUDI) Vrije Universiteit Brussel
Reduction of stochastic noise in physically-based rendering algorithms. KU Leuven
The use of computer graphics has become commonplace in industry, games, films, architectural visualizations, etc. Over the years, there has been an increasing demand for more realism and more graphical fidelity. To achieve this realism, current state-of-the-art illumination algorithms are based on ray tracing, which simulates the physical transport of light through three-dimensional scenes. While ray tracing is conceptually simple, the ...