Home / IT & Computer Science / Biology & Biotechnology / Deep Learning for Bioscientists / Practical: Building a segmentation network
Practical: Building a segmentation network
An article linking to a Colab notebook which demonstrates image segmentation using PyTorch on the Oxford flower dataset.
© The University of Nottingham
An example of using deep learning with PyTorch for image segmentation.
In the linked Colab notebook we will again use the Oxford flower dataset and PyTorch to make a deep learning network that segments flowers within images. We’re able to do this since the dataset comes with pre-segmented versions of the images we can use to train the network.
In the notebook we will cover the following:
- downloading and using the segmented image maps in a new custom Dataset object
- importing and adapting a network from PyTorch we can train to perform the segmentation
- training the network using the binary cross entropy loss function
- testing and visualising the output.
Follow the link below and work through the Colab notebook step by step.
Image segmentation using deep learning in PyTorch
You may wish to open the link in a new browser tab so you can refer back here quickly.
Please leave any questions or comments below.
