Face Recognition App Yielded 80% Accuracy In Telangana Municipal Polls

The Telangana State Election Commission used a first-of-its-kind facial recognition app to counter the impersonation of voters. The app used in the recently concluded municipal polls in Telangana yielded an accuracy of 80 percent, officials have claimed.

"The result of the face recognition app was encouraging. The accuracy was up to 80 percent. Some factors (poor quality of lighting in some polling booths and old pictures) have impacted the accuracy at some places." a senior official said. Otherwise, the result would have been above 90 percent.

The app was used on a pilot basis in the selected 10 polling stations of the Kompally Municipality of Medchal Malkajgiri district in the urban local bodies polls on January 22.

The data collected using the app would be deleted after the polls, an official press release had said earlier.

An additional polling officer first verified the identity proof of the voters and then took photographs and uploaded it using the face recognition app installed on the phone provided to him to compare the same with the photographs of the orders of the polling station concerned.

The app displays the results of the verification based on the match established with any one of the voters with an appropriate message.

However, Owaisi objected to using the technology alleging that it was dangerous and would invade privacy

"Use of facial recognition technology is dangerous and will cause irreversible damage to our right to privacy as individuals. There must be a moratorium on its use. Instead, it is seeing widespread use despite SC judgment (Puttaswamy)," Owaisi tweeted on Tuesday reacting to a media report.