A deep learning app, Chakshu Face Attendance is created by Kartikeya Bhardwaj, a mechanical engineering student in the class of 2020 at Mahindra’s Ecole Centrale School of Engineering. This is to register a touchless recording of attendance for organisations, institutions, and employers. The app is particularly useful in the times of Covid-19 outbreak and provides a safer alternative as compared to finger-touch based attendance systems. The start-up is being incubated at Mahindra e-hub.     

Speaking about encouraging entrepreneurial efforts, Dr. Yajulu Medury, Vice Chancellor Mahindra University says, “One of our focuses has been on Artificial Intelligence and Deep Learning for which we have not just created state-of-the-art labs but also made available live projects for students to work on. Our Centre for Entrepreneurship and Innovation provides all necessary assistance in sync with the Government of India’s mission”, says Dr. Yajulu Medury, Vice Chancellor, Mahindra University.

Fingerprint or ID scanners are a norm for attendance in big organisations. However, MSMEs refrain from it because of exorbitant costs involved in installing RFID systems. Furthermore, such systems require regular maintenance and expensive hardware. Other constraint of using it involves proxy attendance in which one employee punches the attendance for other one. This results in heavy productivity losses for any organisation.    

Hence, Chakshu.ai has come up with an app named Chakshu Face Attendance. It is a touch less app that helps an organisation in clocking attendance through face recognition with the help of an android device. The SpoofSense feature in it effectively handles buddy punching as Chakshu’s AI algorithm detects Spoofing attacks.

The app uses PyTorch (a Powerful Machine Learning Framework by Facebook) to implement Deep learning algorithms. Involving AI algorithms to thousands of users usually results in increased latency situations. But, Chakshu App involves carefully built model optimisation that enables the backend to handle huge volume of traffic. This further involves a frictionless user experience.