AutoEncoder_LArTPC(Keras): AutoEncoder for LArTPC, Keras version

AutoEncoder applicaiton for LArTPC. Network taks a 512x512 LArTPC images and returns score value for weather it contains a neutrion of just cosmics.

Dependecies:

LArCV, ROOT, Keras

Setup:

0. Setup LArCV using Wiki,
1. git clone https://github.com/ruian1/autoencoder_lartpc.git
2. source setup.sh

Training:

0. Edit training configures under ./uboone/
1. python ./uboone/train_autoencoder_2.py ./uboone/autoencoder_2.cfg

Inference:

0. cd ./uboone/production/
1. python inference_mrcnn_center.py image2d.root (for image2d input).
2. python inference_autoencoder.py pixel2d.root (for pixel2d input).