Accurate 3D human skeleton reconstruction is essential for applications like human–robot interaction, AR/VR, and motion analysis. In this project, you are going to create ground truth of 3D human skeletons from RGB-D sensors. Tasks: 1. Fusion of RGB-based 2D keypoints (obtained from mmpose) with depth information to reconstruct 3D human joint coordinates. 2. Post-processing of the reconstructed skeletons by enforcing human physical and kinematic constraints. 3. Evaluating the reconstructed skeletons by manual refinement to produce final labels.
| Frequency | Weekday | Time | Format / Place | Period | |
|---|---|---|---|---|---|
| by appointment | n.V. | CITEC 0.114 | 13.04.-24.07.2026 |
| Module | Course | Requirements | |
|---|---|---|---|
| 39-Inf-WP-CIT-x Cognitive Interaction Technology (Focus) Kognitive Interaktionstechnologie (Schwerpunkt) | Vertiefendes Projekt | Student information | |
| - | Graded examination | Student information | |
| 39-Inf-WP-KI-x Artificial Intelligence (Focus) Künstliche Intelligenz (Schwerpunkt) | Vertiefendes Projekt | Student information | |
| - | Graded examination | Student information |
The binding module descriptions contain further information, including specifications on the "types of assignments" students need to complete. In cases where a module description mentions more than one kind of assignment, the respective member of the teaching staff will decide which task(s) they assign the students.