- We make an instance segmentation model that will collect character portraits and 2D sprites to speed up data collection process.
- img2img model to generate 2D sprites from character portraits. Start with
idle-downward
generation, and then expand to creating whole set of sprites.
python -m venv venv
# windows powershell
$ . venv/Scripts/activate
# macOS/Linux
$ . venv/bin/activate
# setup environment
$ pip install -e .
# start label studio
$ label-studio start
# split YOLO
python -m scripts.split_yolo_exported_files --input-dir .\data\instance_segmentation_yolo --output-dir .\data\instance_segmentation_yolo