Automated semantic segmentation of objects in pictures and video sequences
The aim of this project is the development and establishment of automated semantic segmentation of objects in pictures, video sequences and the automated replacement of background images matching different motives.
The need in the industry is to separate different objects in the video sequences from the original background and to position them to a different background. The project would be particularly interesting for the Industry 4.0 sector, where processes are increasingly automated and secured using deep and machine learning methods.
We have experience in deep and machine learning, segmentation, image and video processing.
We are looking for research partners (a researcher or a team) who can support us technically in the creation of the universal Deep Learning model. This model must be able to segment different types of objects with a sufficiently high accuracy.