Research Interest in Computer Vision
- Understanding of multi-view image content
- Semantic enrichment, processing and modeling of xD data
- Efficient scene representations
- Modality fusion
- Geometry estimation from real world images
- Level of detail representations for geometry as well as for semantics
During my PhD work, I mainly focused on efficient representations to combine 2D and 3D data sources
for various computer visions tasks like rapid pixel-wise image classification or object detection.
Further I investigated how redundant input data can be exploited to obtain improved results.
For more details please visit my former research homepage
There you can also find implementations
of the Sigma Points descriptor as well as links to papers.