U-Net: Convolutional Networks for Biomedical Image Segmentation

Betreuerin:
Rania Rayyes

Convolutional Neural Networks (CNN) have demonstrated good results for cancer detection. One of the most promising CNN architectures for efficient and fast tumor detection from small data sets is the well known U-Net architecture. In this seminar topic, the student need to understand and describe the UNet architecture with Conv and Deconv blocks. The student also need to search and present other Padding technique.

References:
https://link.springer.com/chapter/10.1007/978-3-319-24574-4_28
https://lmb.informatik.uni-freiburg.de/people/ronneber/u-net/