New, more precise 3D reconstructions for autonomous driving and preservation of cultural artefacts
From two images to a 3D object
In recent years, neural methods have become widespread in camera-based reconstructions. In most cases, however, hundreds of camera perspectives are needed. Meanwhile, conventional photometric methods exist which can compute highly precise reconstructions even from objects with textureless surfaces. However, these typically work only under controlled lab conditions.
More precise reconstructions despite small numbers of data points
Daniel Cremers, professor of Computer Vision and Artificial Intelligence at TUM and leader of the Munich Center for Machine Learning (MCML) and a director of the Munich Data Science Institute (MDSI) has developed a method together with his team that utilizes the two approaches. It combines a neural network of the surface with a precise model of the illumination process that considers the light absorption and the distance between the object and the light source. The brightness in the images is used to determine the angle and distance of the surface relative to the light source. “That enables us to model the objects with much greater precision than existing processes. We can use the natural surroundings and can reconstruct relatively textureless objects for our reconstructions,” says Daniel Cremers.
Applications in autonomous driving and preservation of historical artefacts
The method can be used to preserve historical monuments or digitize museum exhibits. If these are destroyed or decay over time, photographic images can be used to reconstruct the originals and create authentic replicas. The team of Prof. Cremers also develops neural camera-based reconstruction methods for autonomous driving, where a camera films the vehicle's surroundings. The autonomous car can model its surroundings in real-time, develop a three-dimensional representation of the scene, and use it to make decisions. The process is based on neural networks that predict 3D point clouds for individual video images that are then merged into a large-scale model of the roads travelled.
Artificial Intelligence
AI applications are already changing the world of work, research and everyday life. We are helping to shape this technological progress: Our researchers are developing new methods and applications of AI across disciplines.
Mohammed Brahimi, Bjoern Haefner, Zhenzhang Ye, Bastian Goldluecke, Daniel Cremers. Sparse Views, Near Light: A Practical Paradigm for Uncalibrated Point-light Photometric Stereo. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2024, pp. 11862-11872.
- Daniel Cremers is a professor of computer vision and artificial intelligence. He is also a director of the Munich Data Science Institute (MDSI) and the Munich Center for Machine Learning (MCML). His research is focused mainly on computer vision and machine learning with the goal of developing algorithms for precise 3D reconstructions and image analysis.
- The study will be presented at the Conference on Computer Vision and Pattern Recognition (CVPR) in Seattle (USA) from June 17th to 21st, 2024. The CVPR is the largest and most important event in its field.
Technical University of Munich
Corporate Communications Center
- Julia Rinner
- presse@tum.de
- Teamwebsite
Contacts to this article:
Daniel Cremers
Technical University of Munich
Professorship for Computer Vision and Artificial Intelligence
cremers@tum.de
