Police in Hyderabad, over the past few weeks, have identified around 10,000 cases of COVID-19 mask violations through Closed-Circuit Television (CCTV) cameras installed across the city. The three commissionerates — Hyderabad, Cyberabad and Rachakonda — in the city have ramped up their efforts to fine mask violators through the use of artificial intelligence (AI) tools, as COVID-19 cases in the city began to rise in the last month or so.
There are 375,000 lakh cameras fixed within the Hyderabad district limits and 650,000 cameras installed in the Hyderabad, Cyberabad and the Rachakonda Commissionerates collectively, according to Telangana Home Minister Mohammed Mahmood Ali. However, M Narsing Rao in the IT team of Hyderabad City Police told MediaNama that around 2,000 CCTV cameras have been enabled with artificial intelligence under the Greater Hyderabad area to monitor mask violators.
Narsing Rao said, “If all the cameras are enabled to track mask violators, then in a day around 50,000 to 60,000 challans may be generated — which is not good. Our focus ( by limiting AI to 2,000 cameras) is just to create awareness. If a case is booked against someone, automatically the other person will wear a mask.” It is important to note that this number is only for the Greater Hyderabad limits. Earlier, in 2020, across the three commissonerates, only 335 cameras were AI enabled.
Over a month ago, the Telangana government issued orders directing citizens to wear masks in public places in view of the rising cases in the country. As a result, the city police has deployed artificial intelligence tools as part of its CCTV network to identify mask violators.
In Cyberabad alone, the police have identified over 6,000 mask violators in the last week or so, Cyberabad Police Comimissioner VC Sajjanar told MediaNama. He said that across Telangana there may be around 24,000 cases that have been registered through this system. “We are able to do it through facial recognition and various artificial intelligence algorithms. We have integrated it through the e-challan system.” he said.
According to a recent report by Telangana Today, this action was preceded by strict directions by Director-General of Police M Mahendar Reddy to Police Commissioners and Superintendents of Police (SPs) in districts to enforce mask rule. Explaining the way the process worked, the report said that surveillance cameras tracked down those not wearing masks and mobile patrols reached to the spots to ensure they wear them, apart from penalising them.
Human cost of unaccountable tech
Faraz (name changed), a driver working for a prominent social activist and politician in Hyderabad, used to frequent out of the city regularly during the lockdown last year to take the activist to his destination. Sometimes during these trips he used to forget wearing his mask or helmet.This was at a time when the Telangana government had declared not wearing masks was an offence and also deployed AI-enabled CCTVs to spot such violations.
Usually whenever the police or the technology spot such offences, the system creates an e-challan ticket and the user is notified through a text message on their phones.However a nonchalant Faraz went about his daily routine as he did not receive any such alerts.
“One day he got a court order saying that he has been fined of Rs 8,700. The reasons were myriad: mask violation, not wearing a helmet and so on. But he was never alerted of these challans individually,” said Amjed Ullah Khan, a spokesperson for the Majlis Bachao Tehreek, while talking about the individual who works as his driver.
Khan claimed there were many during the 2020 lockdown who were fined in similar manner — spotted through CCTVs or photo captured by police personnel — and were not informed of their offence immediately. The fines accumulated and many like the driver only came to know of the situation when a court notice was delivered at their residence.
Similarly, a month ago,Kalam (name changed) was fined near the Abids X Roads for not wearing a mask. However, he was not sure whether the image that was captured to show he was not wearing a mask was from a CCTV or a digital camera operated by a traffic police personnel.
First taken up in 2020 in Hyderabad
In May 2020, at the peak of the firstwave of COVID-19, the DGP announced that the CCTV cameras would be integrated with AI tools to fine mask violators across the city.
#AI based #FaceMaskViolationEnforcement is being rolled out by TS police.
Leveraging ComputerVision & #DeepLearningTechnique being implemented on surveillance CCTVs across the cities is #FirstOfItsKind in INDIA.
Shall be enabled shortly across the 3Commissionerates
— DGP TELANGANA POLICE (@TelanganaDGP) May 8, 2020
In the following months, several other States adopted the technology. Madurai City Police in Tamil Nadu rolled out the CCTV-mask violation technology in November on a trial basis at crowded places, a report in the Hindu said. Gujarat police too resorted to similar method to identify people not wearing masks. Around 1.5 lakh persons in Rajkot were identified using CCTV not wearing masks in the past one year, as per an Indian Express report.
‘Nothing but criminalising the poor’
Concerns regarding this move was first highlighted by several activists when it was first introduced last year. Many questioned the security of the servers where this information was being stored, and many raised concerns regarding privacy.
Social activist SQ Masood told MediaNama that apart from privacy concerns, this is just criminalisation of the poor and regular citizens. “To book an offence based on CCTV footage is a very arbitrary method. This leaves no room for discussion which would otherwise be possible if it were in person. If it were done manually, the citizen could present his case to the police,” he said.
- Big brother is watching: The right to privacy for minors
- Exclusive: Hyderabad Police wants to acquire Cellebrite UFED to break into smartphones, extract WhatsApp data
- All police stations must be equipped with CCTV cameras, persons in custody have right to secure CCTV footage: SC
- Face recognition enabled CCTV surveillance systems monitoring quarantine wards in Dehradun
***Correction (7:54 PM, April 20): Added inputs from B Narsing Rao, inspector, Hyderabad City Police. Changed headline; earlier version implied that 650,000 CCTV cameras were AI-enabled, this was incorrect. Originally published at 5.46 PM, April 20.