ETH Zürich is one of the world’s leading universities for science and technology and a driving force behind European excellence in AI and robotics. Its Computer Vision and Geometry Group (CVG), led by Prof. Marc Pollefeys, is internationally recognised for research at the intersection of computer vision, machine learning and robotics. CVG pioneered automated 3D reconstruction from images and develops the methods that allow machines to perceive, map and understand the 3D world, from SLAM and visual localization to semantic, open-vocabulary 3D scene understanding. Its algorithms, open-source tools and benchmarks are used across academia and industry worldwide.
Role in GRAIL
ETH Zürich’s Computer Vision and Geometry Group provides the spatial perception and 3D scene understanding that ground GRAIL’s foundation models in the physical world. It contributes visual localization, mapping and open-vocabulary scene representations that connect GRAIL’s planning and language outputs to real environments across robot embodiments. CVG further develops the benchmarks and evaluation protocols for spatial grounding, anchoring GRAIL’s scientific excellence in Europe.
