CI-Net

This repository contains the CI-Net processing, training, inference, and validation code.

Directory structure

CI-Net/
├── code/
│   ├── data_preparing/
│   ├── labeling/
│   ├── final_preprocess/
│   ├── training/
│   └── validation/
├── raw_data/
└── result/

The directories under code follow the processing order:

1. data_preparing: read and align satellite and radar inputs.
2. labeling: create cloud labels.
3. final_preprocess: convert the prepared fields into model input files.
4. training: train the model and run inference.
5. validation: create validation targets and calculate validation metrics.

Each code directory contains Python files in src and its configuration and shell entry points in run.

raw_data contains the input example included with this repository. result contains prepared data, model files, inference outputs, and validation outputs. See result/README.md for the meaning of its subdirectories.

Environment

conda env create -f environment.yml
conda activate ci-net

Usage

The example preprocessing stages can be run in order:

code/data_preparing/run/run.sh
code/labeling/run/run.sh
code/final_preprocess/run/run.sh

Training and inference use separate configuration files and entry points:

code/training/run/train.sh
code/training/run/inference.sh

Validation target creation and object validation are also separate:

code/validation/run/targets.sh
code/validation/run/validation.sh

The YAML file beside each shell script controls its input and output paths. The shell scripts locate the repository root automatically.

License

The model weights and distributed data under raw_data/ and result/ are licensed under the Creative Commons Attribution 4.0 International License (CC BY 4.0), as described in LICENSE_MODEL_DATA.

The source code under code/ is licensed under the MIT License, as described in LICENSE_CODE.