MASKED FACE RECOGNITION DATASET AND APPLICATION
Researchers from Wuhan University have recently claimed they have developed facial recognition software which can verify the identities of people wearing face masks. In a paper released by the University named – MASKED FACE RECOGNITION DATASET AND APPLICATION -the team said the software boasts 95 percent accuracy rate.
Currently, the available facial recognition devices cannot identify individuals with face masks. The new software aims to solve this problem as face masks are a regular feature now due to coronavirus epidemic.
The research was carried out by Wang Zhongyuan, a professor of computer science at Wuhan University. He was supported by a team of over 10 students who worked remotely from their homes during the lockdown.
WHAT IS FACIAL RECOGNITION TECHNOLOGY?
Facial recognition technology uses a database of photos to identify people in security photos and videos.
Essentially, it uses biometrics to map facial features and help verify identity through key features of the face.
The key feature is the geometry of a face. For example, the geometry of a face is the distance between a person’s eyes and the distance from their forehead to their chin.
Once this geometry is identified, it creates a ‘signature’, also known as “facial signature.”
It is a mathematical formula that is then compared to a database of known faces.
HOW ARE YOU UNKNOWINGLY CONTRIBUTING TO FACIAL RECOGNITION?
In 2010 Facebook began using facial recognition when it automatically tagged people in photos using its tag suggestions tool. The tool scans a user’s face. It then offers suggestions about who that person is.
When you go on tagging you best friend on Facebook or other social media, you are actually unknowingly helping the Artificial Intelligence in “deep learning” about that particular friend you are tagging; and of course about yourself too.
Facial recognition technology is a common feature now. It’s there on Facebook, tagging photos from university campus, your friend’s wedding and the evening parties with friends. Google, Microsoft, Apple and others have built it into apps to compile albums of people who hang out together.
THE INTERNET HOLDS BILLIONS OF PHOTOS
Big data, deep convolutional neural networks and powerful graphics processing units (GPUs) have played a major part in fostering facial recognition technology.
Whether we should be worried or not about our privacy, but thanks to Flickr, Instagram, Facebook, Google and others, the internet holds billions of photos of people’s faces. These have been scraped together into massive image datasets.
They are used to train deep neural networks to detect and recognise faces. GPUs, the superfast chips that are dedicated to graphics processing, do the computational work.
Over the past decade in particular, facial recognition systems have been deployed at several place and the data gathered from them has helped companies hone their technology.
SHOULD YOU BE CAREFUL
Yes, your face mask selfies are seen by your friends and family. But since they are on the public domain, they could be easily collected by researchers looking to use them to improve facial recognition algorithms.
The COVID-19 situation has seen people wearing face masks in public places. It is like an opportunity for facial recognition companies. The more photos they have, the easier it is will be to train their algorithms. They ‘love’ these photos. But, should you be worried?