Repository used to Train and Test Neural Nets for Autism Spectrum Disorder (detects Therapist, Toddler and Teddy Bear)
- Clone YOLOv5 and annotate its path.
- Clone this repo and install its requirements.
- Update YOLOv5
path
variable in the config file.
- Update YOLOv5
# Clone YOLOv5 and store its home location since it is not pip installable
git clone https://github.com/ultralytics/yolov5.git
# Clone and install this repo
git clone https://github.com/ASDDataMining/therapy-aid-nn.git
cd therapy-aid-nn
# python3 -m venv venv; source venv/bin/activate # optional: python virtual environment
pip install -r requirements.txt
# Then go to cfg/yolo_location.cfg and change yolov5's path
- After installation you can adjust the configuration files to your needs and run any of the corresponding
.py
files:train.py
anddetect.py
. - If wandb was installed correctly you can make logs in it
- For downloading labelbox images with the script provided you must create a
.env
file at the repo top-level and editLABELBOX_API=<your-labelbox-api>
.