How Using Facial Recognition & AI May Result In Better Security
Updated: Jul 21
Nowadays, it is very typical to identify a defaulter or criminal by reviewing the images and visual feeds from CCTVs installed in private and public premises. While this may seem to be a sufficient solution to the rising crime rates all over the world, it actually isn’t.
Abhijit Shanbahg, CEO of Graymatics, discusses how facial recognition and Video AI technology is being utilised to protect people, society, and communities in Singapore.
With the growth of facial recognition capabilities and video AI technology, improvements in security and surveillance are no longer bound by historical CCTV feeds and large numbers of security personnel. Imagine if the CCTVs installed in your private condominium complexes can identify the defaulter’s facial characteristics and can trigger an alarm to the guard house when they enter a forbidden area or try to vandalise property. The security guard can immediately rush to catch the defaulter. The possibilities of using AI and facial recognition to maintain security are endless.
AI is truly transforming the landscape of the surveillance industry. CCTV monitoring is only as accurate and reliable as the person monitoring it on the screen. Being humans, operators tend to miss out crucial security clues that can lead to disastrous consequences. AI can be trusted to be accurate at all times and make up for human deficiency. While there is the facility to capture high-resolution images using ultra-high definition cameras, their usage as part of video surveillance solutions remains low. Artificial intelligence can be used to sharpen low quality images to enable security personnel to draw meaningful information out of them.
This is possible with our state-of-the-art facial detection technology that can track emotions and facial features. The first line of defence to improve security at any location, especially offices and manufacturing sites is to only allow staff members access into the premises. Using deep learning and facial recognition, the solution helps companies grant automatic access to their staff and triggers alerts when there is a visitor. The most essential feature of emotion detection enables the solution to capture potential violence or abuse, which is simply not possible through regular CCTV feeds. Smart CCTVs are gradually gaining popularity in Singapore. To harness their social impact and use them responsibly, the National University of Singapore has installed multiple smart CCTV cameras that can detect violence and abuse to increase student and campus safety.
According to the University’s Associate Professor Sim, the technology’s “facial recognition software works by using face detection to locate the exact position and size of a face in an image captured by a camera”. Apart from schools and universities, face detection can be used to manage attendance and access control in large corporate offices, ports, manufacturing sites, retail stores and malls where it is imperative that only employees access certain areas.
In the event of a burglary or break-in, the CCTV feeds can capture the facial characteristics of the burglar and can trigger alerts when he or she is seen in other locations. Recently, Singapore police had announced that they would be installing thousands of CCTVs in Singapore to leverage video analytics technology to track vehicles through object detection. In public areas, with improved federated learning technology, these cameras can also harness the benefits of facial detection to monitor crimes.
The new technology has the advantage of not requiring physical contact, unlike fingerprint recognition. They are using an artificial intelligence technique called machine learning, with features such as the location of the eyes and the shapes of the jaw, lips and nose – are matched against a database.
AI and video analytics go beyond the limitations of manual security monitoring. Graymatics’ Smart Surveillance technology uses AI and image recognition in order to scan spaces, while the surveillance technology services offered include People and Clothing analytics, whereby the technology can pick up on human demographics and physical attributes. Our cutting edge technology can identify and trigger emergency alerts during events when security is compromised like vandalism, break-ins or burglaries, loitering, intruding in forbidden areas and physical fights through movement detection.
These solutions are becoming increasingly deployed across various entities, from governments and large real estates to retail stores, smart cities to commercial banks. There is an increasing need to reduce dependence on humans for security and surveillance as it will be subject to bias, oversight and misinterpretation. It is important to understand this requirement and embed security and surveillance features across technology, so that it is correctly programmed to detect all intrusions and monitor blacklisted people on a seamless, 24/7 basis.
Written by Abhijit Shanbahg, CEO of Graymatics