Image Recognition as a Service

The aim of this project was to develop an auto scalable cloud application which will provide Real Time Object Detection as a Service to users by using AWS cloud resources to perform image detection using the darknet machine learning model on real time videos generated by the Raspberry Pi. This application invokes the machine learning model on this video and detects the objects present in this video. It handles the multiple requests concurrently and will automatically scale out whenever the number of requests increases, that means the number of cloud resources being allocated to the application increases automatically and scales in when the demand reduces. The cloud resources used in this project are AWS Elastic Compute Cloud , Simple Storage Service, Simple Queue Service.