Zurich University of Applied Sciences

Open Opportunities

3D image-based Completion network (Semantic Scene Completion from dual Depth Images) Zurich University of Applied Sciences Institute of Mechatronic Systems 3D shape completion plays an important role in robotics and perception and has obtained significant development in recent years. Existing methods have shown impressive performance on various formats: voxel grids, meshes and point clouds. Object completion focus on the reconstruction of missing topologies and geometries of objects based on partial and incomplete observations coming from one or more 2D images and 3D depth information, for example, point clouds captured against occlusions, under weak illumination, or from limited viewpoints. Computer Vision, Image Processing, Intelligent Robotics, Neural Networks, Genetic Alogrithms and Fuzzy Logic Bachelor Thesis, Master Thesis

Real-time Point cloud based human position estimation and tracking Zurich University of Applied Sciences Institute of Mechatronic Systems Real-time pose estimation and tracking of humans is essential for many applications, and combining multiple sensors in particular vision based sensors increases opportunities to improve human motion tracking. Full body 3D pose reconstruction from single view images is difficult and suffers from the lack of 3D information, compared with two-dimensional or 3D pose estimation from multiple views. It is also very problematic to work with the image plane data since different body parts appear in different sizes and may be occluded depending on how the body is positioned relative to the camera. Artificial Intelligence and Signal and Image Processing Bachelor Thesis, Master Thesis