How 188 inmates in Sharjah jail got help after cameras with AI detect suspicious behaviour
Cameras in cells, corridors help make inferences from facial expressions, body movements
Sharjah: Thanks to a new system that can detect if an inmate has certain behaviourial dispositions, as many as 188 inmates in the Sharjah Punitive and Rehabilitation Establishment have been able to receive timely help.
A recently introduced smart detector analyses footage from surveillance cameras installed in cells and corridors, detecting signs of likely violent or unnatural behaviour, following which authorities are duly alerted.
Lieutenant Sultan Kanon Al Shamsi, Director of the Strategy and Performance Development Branch in Sharjah Punitive and Rehabilitation Establishment, told Gulf News: “The facial expressions and unnatural behaviour of 188 inmates from different nationalities, including both males and females, were captured by the system, following which they were moniotred for some time and then summoned by the administration staff. Upon analysis of the signs, the system issued red alerts in applicable cases on the screen of operation room. The inmates who suffered from a psychological problems were thus able to inmate was be identified and get the help they needed.”
In one case, an Asian inmate who was displaying signs of sadness, was summoned by staff. Upon interrogation, he revealed that his father was sick and was admitted in hospital. Neither he nor his family were able to pay hospital bill. Understanding his plight, the facility helped the inmate and arranged to pay his father’s bill in cooperation with a charity.
In another case, another depressed inmate confessed that the school fees of his children were unpaid. The facility again arranged to pay up the dues with the help of a charity.
Besides depression, even anger was managed with the help of the smart system. Lieutenant Al Shamsi said, “The system helped in reduced fighting among inmates by 20 per cent. No suicide or attempt to escape was recorded in the facility.”
Brigadier Ahmad Abdul Aziz Shuhail, Director-General of the Sharjah Punitive and Rehabilitation Establishment, told Gulf News that the system is based on artificial intelligence (AI) and uses “gait analysis” that allows a computer to analyse human motion and behaviour and thereby pre-empt any unusual act.
The system is linked to a programme of surveillance cameras that monitors prisoners round the clock. The facial expressions, moods, movement of the hands and other parts of the body can be analysed and a warning be sent out about a prisoner even before he or she can commit an untoward act. The smart monitoring system uses AI and machine-learning algorithms, the official explained.
Brigadier Shuhail pointed out that the security system can identify anger, joy, sadness, frustration and other emotions and predict likely actions, using big data and smart technologies. It then begins an automated analysis and issues an alert to the staff about possible scenarios.
What is gait analysis?
Gait analysis (as it is commonly called) involves evaluation of gait (or the manner in which a person walks) and based on that evaluation, the system can preempt any likely untoward action by an inmate. This analysis is based on video footage obtained from the surveillance cameras or closed-circuit televisions installed on the prison premises.
The video analytics software not only detects suspicious behaviour, but also detects items such as phones, drugs or weapons that can potentially be used with a harmful intent.