Loss functions benchmark for brain tumour segmentation
From LRDE
The printable version is no longer supported and may have rendering errors. Please update your browser bookmarks and please use the default browser print function instead.
- Authors
- Lukas Rabier
- Type
- techreport
- Year
- 2020
- Number
- 2015
Abstract
Training a convolutional neural network relies on the use of loss functions, which provide an evaluation of the performance that allows the network's optimisation. Different loss functions evaluate performance in different ways, and thus affect the training differently. This report aims to share our progress in evaluating the performance of various loss functions in the training of convolutional neural networks for brain tumour segmentation.