Skip to content

Latest commit

 

History

History
56 lines (50 loc) · 2.7 KB

README.md

File metadata and controls

56 lines (50 loc) · 2.7 KB

MethaneMapper Training code

MethaneMapper repository structure is similar to DETR repository from META. The training policy involves first training for bounding box detection of methane plumes, followed by fine-tuning for methane plume segmentation mask.

There are no extra compiled components in MethaneMapper and package dependencies are minimal mentioned in the requirements.txt. Codebase is very simple to use.

Steps:

  1. Clone the repository
git clone https://github.com/UCSB-VRL/MethaneMapper-Spectral-Absorption-aware-Hyperspectral-Transformer-for-Methane-Detection.git
  1. Install dependencies
pip install -r requirements.txt
  1. That's all! It should be good to train and evaluate MethaneMapper

Data preparation

Download MHS dataset by following these steps. DATA_DOWNLOAD

Directory structure from download :

./training_dataset2020 OR ./training_dataset16171819
|_ mf_output
|_ mf_output_norm
|_ mf_tiles #matched filter tiles
|_ mf_tiles_no_norm
|_ rdata_tiles #raw 90 bands from hyperspectral image
|_ rgb_tiles #reconstructed RGB image from hyperspectral image

No create soft links to the "./data/train_dataset" folder as follows to avoud duplication of dataset (~size 10 TB):

  1. matched_filter output tiles to mf_tiles folder
ln -s ./training_dataset2020/mf_tiles methanemapper/data/train_dataset/train/mf_tiles
ln -s ./training_dataset2020/mf_tiles methanemapper/data/train_dataset/val/mf_tiles
  1. Link the raw 90 bands tiles to rgb_tiles folder
ln -s ./training_dataset2020/rgb_tiles methanemapper/data/train_dataset/train/rgb_tiles
ln -s ./training_dataset2020/rgb_tiles methanemapper/data/train_dataset/val/rgb_tiles
  1. Link the reconstructed RGB image tiles to rdata_tiles folder
ln -s ./training_dataset2020/rdata_tiles methanemapper/data/train_dataset/train/rdata_tiles
ln -s ./training_dataset2020/rdata_tiles methanemapper/data/train_dataset/val/rdata_tiles

Training methane plume bounding box detector

./plume_box_run.sh

Fine-tuning plume segmentation mask detector

./plume_mask_run.sh