DROPOUT CONCRETE AUTOENCODER FOR BAND SELECTION ON HSI SCENES

PROFESSIONAL README GENERATOR

Acknowledgment: This code is mainly based on the works dl_selection.

I. Training.

./train.sh --cfg configs/xxx.yaml

II. Prepare Data.

The project folder structure should look like this:

|--myHyper
  |-- $configs
  |   |-- KSC_T1.yaml
  |-- $datasets
  |   |-- KSC_gt.mat
  |   |-- KSC.mat
  |-- data.py
  |-- encoders.py
  |-- concreteVAE.py
  |-- utils.py
  |-- train.py
  |-- train.sh