coffee-leaderboard

This repository is part of a blog post that guides users through creating a coffee leaderboard that uses face detection to track the number of coffee people drink using the AWS DeepLens

Following the steps described in the blog post, the final architecture is this:

face_function.py

Using Amazon Rekognition, this lambda function responsible for recognising/registering a face and mug, storing the results in DynamoDB

deeplens_inference_function.py

This lambda function runs on the AWS DeepLens and perform inferences and the necessary logic. It uploads frames to Amazon S3 when a face is detected, as well as adds features such as a cooldown period between uploads along with a countdown before taking a picture.

app/

This folder contains a Python Flask application that presents the information collected. Using AWS Elastic Beanstalk, it is easy to deploy this application and visualise the result collected from the AWS DeepLens