HorizonMS is a deep learning framework developed by PyTorch. It was developed by Medical Imaging Technology Lab (MITL) at Sibionics and was released on December 12th, 2023. Its documentation is here.
HorizonMS is a foundational library for computer vision research and it provides the following functionalities:
-
It supports image classification, image segmentation, and object detection.
-
It supports data augmentation by PyTorch, OpenCV, or mixture of OpenCV and PyTorch
-
It configures the experiments by dictionary.
-
It discriminates softmax, sigmoid, and linear activation for the network output in both losses and metrics.
-
It supports well-known deep learning algorithms.
HorizonMS supports both pip installation and local installation. The command for installations are as follows:
1. pip installation:
pip install horizonms2. local installation:
git clone https://github.com/sibionics-mitl/horizonms.git
cd horizonms
pip install . # method 1, or
python steup.py install # method 2Juan Wang, Xiaochun Pan, Peihai Huang, Wanduo Zheng
@misc{hms,
title={HorizonMS},
author={HorizonMS Contributors},
howpublished = {\url{https://github.com/sibionics-mitl/horizonms}},
year={2023}
}