Gluoncv classification GPU/CPU Inference API

This is a repository for an image classification inference API using the Gluoncv framework.

The inference REST API works on CPU/GPU. It's supported on Windows and Linux Operating systems.

Models trained using our Gluoncv Classification training repository can be deployed in this API. Several models can be loaded and used at the same time.

Prerequisites

• OS:
  • Windows or Linux
• Docker

Check for prerequisites

To check if you have docker-ce installed:

docker --version

Install prerequisites

-Ubuntu

Use the following command to install docker on Ubuntu:

chmod +x install_prerequisites.sh && source install_prerequisites.sh

-Windows 10

To install Docker on Windows, please follow the link.

P.S: For Windows users, open the Docker Desktop menu by clicking the Docker Icon in the Notifications area. Select Settings, and then Advanced tab to adjust the resources available to Docker Engine.

Build The Docker Image

In order to build the project run the following command from the project's root directory:

docker build -t  gluoncv_classification -f {CPU or GPU}/dockerfile .

Behind a proxy

docker build --build-arg http_proxy='' --build-arg https_proxy='' -t gluoncv_classification -f ./{CPU or GPU}/dockerfile .

Run the docker container

To run the API, go the to the API's directory and run the following:

-Using Linux based docker:

• CPU:

sudo docker run -itv $(pwd)/models:/models -p 4343:4343 gluoncv_classification

• GPU:

sudo nvidia-docker run -itv $(pwd)/models:/models -p 4343:4343 gluoncv_classification

-Using Windows based docker:

For windows inference is only supported on CPU

• CPU:

docker run -itv ${PWD}/models:/models -p 4343:4343 gluoncv_classification

The API file will be run automatically, and the service will listen to http requests on the chosen port.

API Endpoints

To see all available endpoints, open your favorite browser and navigate to:

http://localhost:4343/docs

The 'predict_batch' endpoint is not shown on swagger. The list of files input is not yet supported.

P.S: If you are using custom endpoints like /load, /detect, and /get_labels, you should always use the /load endpoint first and then use /detect or /get_labels

Endpoints summary

/load (GET)

Loads all available models and returns every model with it's hashed value. Loaded models are stored and aren't loaded again

/detect (POST)

Performs inference on specified model, image, and returns class

/get_labels (POST)

Returns all of the specified model labels with their hashed values

/models (GET)

Lists all available models

/models/{model_name}/load (GET)

Loads the specified model. Loaded models are stored and aren't loaded again

/models/{model_name}/predict (POST)

Performs inference on specified model, image, and returns class

/models/{model_name}/labels (GET)

Returns all of the specified model labels

/models/{model_name}/config (GET)

Returns the specified model's configuration

Model structure

The folder "models" contains subfolders of all the models to be loaded. Inside each subfolder there should be a:

• classes.txt file: contains the name of the classes separated by a ','

• .params file : contain the models parameters

• -0000.params file : contains the models parameters

• -symbol.json file : contains the models architecture

• Config.json (This is a json file containing information about the model)

     {
         "cpu": true, 
         "max_number_of_predictions": 3, 
         "minimum_confidence": 0.6,
         "inference_engine_name": "classification"
     }
  

  P.S:

  • Make sure to change cpu to {true/false} based on the image used {CPU/GPU}
  • You can change confidence and predictions values while running the API
  • The API will return a response with a confidence higher than the "minimum_confidence" value. A high "minimum_confidence" can show you only accurate predictions
  • The "max_number_of_predictions" value specifies the maximum number of classes returned and analyzed in the API response

Acknowledgements

Roy Anwar,Beirut, Lebanon

Fouad Chaccour, Beirut, Lebanon