TensorFlow Graph Api to execute tensflow based pre-trained models. This project already has 2 models, i.e. Face Detection and Object Detection. This also a TF-Slim model trained using Azure Custom Vision for cloth classification.
This project is a part of Deep Fashion Project and is intended to work along with all other smart micro-services under the project.
Once the all of the requirements are satisfied, we can start playing with the api. In order to get the codebase and execute pre-complied binary, follow the steps:
# install required dependencies
> ./install
# getting the codebase
> git clone [email protected]:retail-in-store-shopper-insights/tf-graph-api.git tfGraphApi
> cd tfGraphApi
# running api
> go run -u objectDetection \ #selecting the base use case
-az clothDetection \ #selecting tf-slim model trained on azure
-img-dir IMAGES \ #input image dir \
-c #remove image from dir after processing