4 "Face Recognition" Myth Busters
Updated: Jun 20, 2022
To boost the post’s SEO, include a keyword in the title.
Myth-1: Face Recognition is expensive, needs a lot of hardware and CCTV cameras
A full fledged Face Recognition is now possible with mobile devices such as phones and tablets. It can be delivered from the cloud, on premise or on the edge. While CCTV is used for use cases where typically for non-cooperative subjects ( such as a retail store where customer’s walk in ), in most cases of Face recognition such as time and attendance or Face verification such as loan applicantion processing etc. can be done on mobile devices.
Myth – 2: Face Recognition needs a strong internet connection to work
While some organizations use Face Recognition plug ins such as AWS , Azure, which require images to go to a central cloud servers to get recognized, FaceTagr’s algorithm performs Face Recognition on the mobile device without any internet, and synchs results to the cloud, whenever 2G level internet is available.
Myth – 3 Face Recognition is not accurate.
Accuracy of Face Recognition depends on a number of factors such as Face box size, pitch, yaw etc. In the case of cooperative faces ( where subjects show their face to the camera ), the accuracy is very high. For example , FaceTagr’s algorithm has a NIST rated accuracy score of 99.55 %.
Myth – 4 Face Recognition is unethical
This idea has been widely discussed in many circles with both pro and nay Sayers. One of the main reasons for the controversy is a model created by a large global player for US law enforcement with a predictive model that indicated a racial bias on a person’s predisposition to commit crime based on face. The other issue is with privacy , as mandated by GDPR on personally identifiable information.
If you would like to know more about us, visit us www. Facetagr.com or email us firstname.lastname@example.org.