Image recognizer web application that runs on Kubernetes with 4 replicas, you can upload your image to get a label and confidence. The underlying model is an AlexNet trained on the famous Imagenet dataset.
- Install Docker
- Install Kubernetes
- Install Git
- Open Docker desktop and run docker daemon.
- Navigate to Kubernetes tab in Docker desktop and select 'Enable Kubernetes' option.
git clone //https://github.com/sushanthsamala/image-recognizer-web-app
cd image-recognizer-web-app
docker build -t image-recognizer .
kubectl apply -f deployment.yaml
- Navigate to http://0.0.0.0:5000 in your browser
- Upload an image.
- The web page displays image label and the confidence of the model on the prediction.
- Repeat step 2 until you are satisfied.
kubectl delete -f deployment.yaml
docker rmi -f $(docker images -a -q)
Instead of build the Dockerfile on your machine, you can directly use my publicly available docker image on docker hub: sushanthsamala/image-recognizer