U-Net: Convolutional Networks
for Biomedical Image
Segmentation
A network and training strategy that relies on the strong use of data
augmentation to use the available annotated samples more efficiently
The typical use of convolutional networks is on classification tasks, where
the output to an image is a single class label. However, in many visual tasks,
especially in biomedical image processing, the desired output should
include localization, i.e., a class label is supposed to be assigned to each
pixel.
U-net architecture. Each blue box corresponds to a multi-channel feature map. The number
of channels is denoted on the top of the box. White boxes represent copied features maps.
Overlap-tile strategy: Allows deep learning models to efficiently process large
images by dividing them into smaller overlapping tiles, ensuring that the
networks has sufficient context information for accurate predictions across the
entire image.
U-Net: Convolutional Networks for Biomedical Image Segmentation 1
for Biomedical Image
Segmentation
A network and training strategy that relies on the strong use of data
augmentation to use the available annotated samples more efficiently
The typical use of convolutional networks is on classification tasks, where
the output to an image is a single class label. However, in many visual tasks,
especially in biomedical image processing, the desired output should
include localization, i.e., a class label is supposed to be assigned to each
pixel.
U-net architecture. Each blue box corresponds to a multi-channel feature map. The number
of channels is denoted on the top of the box. White boxes represent copied features maps.
Overlap-tile strategy: Allows deep learning models to efficiently process large
images by dividing them into smaller overlapping tiles, ensuring that the
networks has sufficient context information for accurate predictions across the
entire image.
U-Net: Convolutional Networks for Biomedical Image Segmentation 1
