tuhoag/classification-mxnet

Training Scripts for MXNet

I implement this project to bring Tensorlow and PyTorch experience to those who are first to MXNet. The script uses a state-of-the-art deep learning model (Resnet) to train classifiers for images. The script is used to detect markers in X-Ray images and the experimental results are published in an academia paper[1]. The full source code of the project is confidential so I only share the training script.

[1] V. Nguyen, J. De Beenhouwer, S. Bazrafkan, A-T Hoang, S. Van Wassenbergh, and J. Sijbers. 2020. BeadNet: A Network for Automated Spherical Marker Detection in Radiographs for Geometry Calibration. CTMeeting-2020, Regensburg, Germany.