Single shot - Face verification System
A single-shot face detection system using state-of-the-art algorithms to add an additional layer of security to private data and preventing data dilution by implementing user restrictions.
Created on 21st June 2020
•
Single shot - Face verification System
A single-shot face detection system using state-of-the-art algorithms to add an additional layer of security to private data and preventing data dilution by implementing user restrictions.
The problem Single shot - Face verification System solves
In these difficult times, working from home has been a real challenge to all of us. Verifying the identity of a user and authenticity of his work has become a really challenging feat to achieve.In these difficult times, password protected websites do not seem to be full proof solution to verify the actual presence of the desired user.With this project,we aim to tackle data dilution which can be done if a user tries to create multiple accounts and dilute the poll results by those accounts.This project uses a user database which collects the username, password as well as a photo of the user and allows the user to sign in only when his live facecam image, username and password matches with that in the database.Through this we maintain the integrity of data and aim to provide a platform which can help us create awareness about important issues and help people stay updated and aware of the current situation.
Challenges we ran into
Challenges include:
Integrating a deep learning model to the website was a very challenging task and required a lot of time and attention.
We just wanted to decrease the bandwidth and keep the privacy of the user (even while we were taking the image of the user from time to time) , so for that we had to convert the image to some vector before storing and transmitting it.
We tried to make the poll anonymous thus not revealing the name of the user who created any poll or vote any of the poll.
Technologies used