FCN-8S-DATA

Description

BO-FCN is the code accompanying the paper "Automatic identification and application of organic microcomponents in oil shale based on deep learning models using Bayesian optimization algorithm." It combines a Bayesian optimization algorithm with a semantic segmentation model to identify organic microcomponents in oil shale. To use it, first copy the VOC-format dataset together with the training and testing partition files into the dataset folder. Then run train.py to train the model and generate the weight files, run test.py to set the basic parameters and let the Bayesian optimization module test the trained model on the prepared test dataset and output the results, and finally run eval.py to validate the model on the validation dataset and output the validation data and images.

Institutions

• Jilin University

  Jilin, Changchun

Categories

Licence