Simplify objection detection, with robust Web and Telegram bot-based application

About


Detect-me is an object detection application, based on yolo5 algorithm (You Only Look Once). It is a robust and resilient application built with cloud native technologies, to meet the requirement of a modern-day applications.

Solutions


Based on Flask web server, with integration with mongodb microservice along with yolo5 microservice and AWS S3 bucket. In the Web UI, clients can upload their images, and get list of objects detected by Yolov5 microservice service.

Based on telebot python module, with integration with yolo5 microservice and AWS S3 bucket. In Telegram bot, it provides a chat-based interface with directive messages, along with list of objects detected by backend service.

Technologies Used

Contributors

Ashutosh Kapoor

Frontend Developer

Flutter and Web developer. Pursuing BTech CSE with spl. OSS from UPES.
Ashutosh Developed the entire frontend of the project.

Deepanshu Rawat

DevOps and Cloud Engineer

DevOps and Web Enthusiast. Pursuing BTech CSE(Hons.) with spl. DevOps from UPES. Deepanshu's role was to setup services and tools in production in K8S along with Continuous Deployment of app.

Aryansh Rana

DevOps Engineer

Devops and Blockchain enthusiast pursuing BTech CSE(Hons.) with spl. DevOps from UPES. Aryansh's role was to setup a CI / CD pipeline of application using GitHub Actions and Argo CD