20.07.2023ChatGPT but also by image generators like Midjourney and DALL-E. Today, anyone can give simple commands to these programs and obtain images and illustrations that closely resemble real photographs of people and places. Although there are concerns about the dangers of rapid AI development, the explosion of interest in AI offers hope for further advanced research.
“We have established a new research group at IDEAS NCBR dedicated to computer graphics, one of the most significant areas of AI-related research. We have convinced Przemysław Musialski, an outstanding specialist from the New Jersey Institute of Technology, to return to Poland and lead this group,” says Prof. Piotr Sankowski, the president of IDEAS NCBR.
Przemysław Musialski, the leader of the new research group, is a computer science professor at the New Jersey Institute of Technology. His scientific research includes geometric modeling, geometry processing, computational fabrication, and machine learning, with a specialization in algorithmic solutions for generating digital content.
Before joining IDEAS NCBR, Prof. Musialski led the Computational Fabrication group at the Center for Geometry and Computational Design at the Vienna University of Technology. He has also held academic and research positions at VRVis Vienna, TU Vienna, Arizona State University, and King Abdullah University of Science and Technology. He holds a master’s degree in media systems from the Bauhaus University in Weimar and a Ph.D. in computer science from the Vienna University of Technology.
What will the research group do?
The results of graphics research are mainly used in computer image creation, including computer games, animated films, special effects, virtual and augmented reality, and design software in architecture, engineering and other industries. The newly established group will focus on areas combining graphics and machine learning. Ultimately, the team will consist of a dozen or so scientists.
“All our research will be related to the use of AI in computer graphics. Computer graphics as a scientific field is part of computer science and deals with the creation of algorithms for calculating images and geometry. It is a mathematical field, while graphic design deals with the creative modelling of images,” comments prof. Przemysław Musialski.
“The aim of our group’s research is to combine classical algorithms with modern machine learning algorithms for the purposes of neural rendering, i.e. creating images, geometric neural modelling, i.e. creating the geometry of digital worlds, and neural simulation, i.e. animation. These are topics that are applicable in many industries: gaming, film, but also in fields such as architecture, engineering or medicine. I hope that among Polish scientists, famous across the ocean for their outstanding talents in the field of IT, there will be many willing to develop this area of computer science,” adds Prof. Przemysław Musialski.
The main areas of research of the group led by Przemysław Musialski will be 3D rendering, generative modelling, geometric modelling, animation and simulation.
- In computer graphics, the main goal is to create photorealistic images and videos through a process called “rendering”. Over the years, graphic designers have developed advanced mathematical methods to simulate the propagation of light in the generated images. Recently, the concept of neural rendering has emerged that extends the traditional principles of computer graphics by using elements of computer vision and machine learning. This approach allows you to create realistic, high-quality images from new perspectives. Our research explores this technique of synthesizing high-quality visual content for applications such as film and game production.
- Neural modelling – An important aspect of computer graphics is geometric modelling, geometry processing, and shape understanding. Designing geometric models requires not only technical but also artistic skills. Unfortunately, design software often does not meet the user’s requirements, which forces them to look for other ways to implement the idea. Prof. Musialski’s group will conduct research on new algorithms and work on solutions for generating geometric content using generative machine learning methods. These solutions can be used in the entertainment industry, as well as in computer-aided design, engineering and manufacturing (CAD/CAE/CAM).
- Neural simulation and animation – combines the principles of physical simulation and animation with the capabilities of machine learning. This approach is becoming an essential tool for analyzing complex phenomena in the real world (e.g. climate change or crowd behavior in densely populated areas). Physical simulations provide a safe virtual environment for testing various scenarios, supporting decision-making processes, e.g. in the area of The team will use neural networks to transform these challenges into optimization tasks that can be more data-efficient. The research results of Przemysław Musialski’s team in this area will be useful in the production of visual content, the gaming industry, as well as in urban planning and complex simulations related to security or medicine.
Currently, a team of specialists is being built to conduct research in Przemysław Musialski’s research group. In addition, IDEAS NCBR has research groups dealing with: intelligent algorithms and data structures, system security and data privacy, zero-waste machine learning and computer vision, as well as research teams operating in the field of AI for security, precision forestry and algorithms in autonomous UAVs.
See job and internship offers in the new research group:
Postdoc – Advancing Generative Modeling in Computer Graphics
PhD Student – Neural Rendering: Exploring the Intersection of Machine Learning and Computer Graphics
PhD Student – Advancing Geometric Modeling Using Implicit Neural Representations
Intern – Generative Modeling in Computer Graphics
