“A Smart Automatic Contactless Temperature Checkup and Mask Detection Kiosk using Facial Landmarking and Deep Learning for under $100.
Countries around the Globe are Reopening, living with the Novel Coronavirus is becoming the new way of life. But to Stop the Spread of the Virus we need to separate people having the Coronavirus from the Rest.
According to the CDC, fever is the leading symptom of the Coronavirus with up to 83% of Symptomatic Patients showing some signs of fever. Many Countries are making Temperature Checkups and Masks mandatory for Schools, Colleges, Offices, and other Workplaces.
Currently, Temperature Checkups are done manually using Contactless Thermometer. Manual Checkups can be Inefficient, Impractical (in places with a large footfall), and Risky.
To solve these problems, I have designed a Kiosk that automates the process of Temperature Checkup by using Facial Landmarking & Contactless IR Temperature Sensor and Mask Detection using Deep Learning Neural Network for less than $100.
The use of this Kiosk is not limited to Schools, Colleges, Offices, other Workplaces but can also be used at High-Risk Areas like Hospitals. This Device can also be used at Train Stations, Bus Stops, Airports, etc.
My approach for this project was to build a Streamlined Setup process such that anybody without any Prior Experience of Computer Vision or Deep Learning can use this. This is a fully functioning and ready to use Project. I have made this Project highly customizable by adding code files for each and every stand-alone part and the full version. Thus, you may use any of the parts of the project individually.
Thanks to Amazon AWS Credit, I was able to train my Deep Learning Model for Mask Detection with Good Compute Power using AWS EC2 Instance.
What’s new in v2?
All-new Web-based Dashboard Design compatible with almost all browsers
Self Alignment Feature
Reduced Cost for less than $100 ($150 with display)
BalenaFin Support for large scale Deployments, use of Cellular Services to name a few.
PC Support (Mac/Linux/Windows)