Mask Service for Image segmentation wrapped into fastapi.
The model is taken from this github project.
Check it out to see what the model can do.
Installing the dependencies for the RCNN on your system will be painful.
This repo only support the execution using a docker container.
The container image can be build using the provided docker file.
docker build -t expingo-mask-service .
The build takes a while. Once finished, run
docker run -p 8000:8000 expingo-mask-service
You can inspect and play with the API in swagger
http://localhost:8000/docs
To create a usable development setting I recommend using the docker container and mounting the project into the container.
docker run --rm -it -v /path/to/project/on/your/host:/dev -p 8000:8000 expingo-mask-service bash
Using swagger as describe above give an easy way to tease the api with some requests.