Bengaluru’s Central Prison in Parappana Agrahara, near Electronics City, will soon be equipped with an AI-powered video surveillance system. Karnataka Prisons and Correctional Services Department has recently floated a tender to set up an advanced AI-driven audio-video analytics system to enhance monitoring capabilities within the prison.
Officials said that an AI-powered audio-video analytics system will help monitor illegal activities, violence, suicide attempts, vandalism, crowd analysis, prison breaches, unauthorised access, smuggling of contraband items, and the use of unauthorised mobile phones. “This will enable us to receive notifications of illegal activities on a real-time basis, round the clock,” said an official.
Prisons in Punjab, Uttar Pradesh, and Delhi have also deployed AI to monitor their facilities. This development comes after a viral photo and video of jailed actor Darshan receiving preferential treatment at Parappana Agrahara Central Prison, which led to the suspension of nine prison officials on August 25. On August 29, Darshan, an accused in the Renukaswamy murder case, was transferred to Ballari Central Prison.
Separately, Prisons and Correctional Services Department on August 28 floated a tender for the installation of CCTV cameras (surveillance system) in Central Prison at Ballari, where Darshan is currently lodged. According to tender documents, facial recognition software (FRS) will help integrate photographs of prisoners and will also be capable of recognising changes in appearance due to age or grooming, as well as analysing pre-recorded video feeds.
“A key feature of the system is its alert generation capability. It will promptly notify authorities when a blacklisted individual is detected or if someone is found in a restricted area. The system’s deep learning-based technology will continuously improve its detection and recognition accuracy, adapting to new data and evolving threats,” said an official.
“The AI-powered surveillance system will also include advanced features such as reverse facial recognition, which allows authorities to upload a suspicious visitor’s photo and search through camera footage for a match. This capability will be instrumental in post-event forensic analysis, providing crucial evidence in investigations”, the official said.
“The system can capture audio data of persons of interest and perform both recognition and matching, offering insights into voice samples with varying levels of clarity. This feature will be crucial for identifying individuals based on their voice and generating real-time alerts when suspicious audio activity is detected,” the official added.
“The AI audio analytics component will be capable of performing searches on audio databases using voice samples, regardless of the language spoken. The system’s language-agnostic design ensures that it can handle audio data entry, matching, and searches in multiple languages, including English and Hindi. Furthermore, the system will support gunshot detection, SOS detection through specific phrases, and real-time verification using the AI audio database, all integrated with the prison’s existing CCTV infrastructure,” the official further said.