Without a doubt, the newest generation of security systems is here. With the latest machine learning capabilities powering these surveillance systems, cameras are now able to detect trouble with almost zero human supervision. Businesses customarily use them to spot sketchy individuals. Beyond providing security and safety, though, the more advanced surveillance cameras of today can gain crucial insights into customer types and behavior, enabling enterprises to improve operations and boost sales. As artificial intelligence (AI) is taking surveillance technology to greater heights, there are a number of groundbreaking developments unfolding in the market.
Japanese telecom giant NTT East, along with startup Earth Eyes, is developing a new security camera with built-in AI capabilities. It is designed to help shop owners identify possible shoplifters. Dubbed the AI Guardman, this new surveillance system comes with open source technology created by Carnegie Mellon University, one of the United States’ top technical research universities. It scans the live video feed from the store and analyzes the physical behavior exhibited by the individuals caught on camera. The technology then matches the gathered data to a preexisting database of suspicious behaviors. If it detects something sketchy, an app notifies shopkeepers.
NTT East and Earth Eyes have been developing AI Guardman for the last few years. Andin early 2018, they proudly announced some results from the early tests of the camera. According to the reports, the AI Guardman reduced shoplifting in stores by 40 percent.
The tech developers are aiming to push out 10,000 cameras in the next three years. Though their primary target is to introduce automated surveillance systems to large enterprises, they won’t be excluding small businesses.
On the other hand, NTT East still has a lot of work to do before the cameras are ready to hit the shelves. They’ve yet to address issues related to automated surveillance such as privacy, accuracy, and discrimination.
As of now, AI Guardman continues to make common errors when trying to identify suspicious behavior. For instance, indecisive customers who will pick up an item, return it to the shelf only to pick it up again are falsely flagged. Any business couldn’t afford to be associated with these types of mistakes, which is why the AI Guardman team has their work cut out for them still.
Dubai government agencies are using one of the latest breakthroughs in hi-tech surveillance systems. Thousands of cameras in Dubai relay live video feed to a command center. From there, officials monitor traffic, tourism, and even brick-and-mortar establishments. Using a new AI network, cameras enable various government agencies to detect security breaches and criminal activity.
Assistant Commander-in-Chief for Criminal Investigation Affairs Major-General Khalil Ebrahim Al Mansouri announced that these smart cameras are a new project called Oyoon (or eyes), one of Dubai’s 2021 Vision requirements for developing into a smart city. It was set up to address criminal activities, reduce traffic congestion, accidents, and deaths. Oyoon links all government and private sector security systems into one network, making it easier to detect and even prevent serious crimes.
Artificial Intelligence powers these surveillance cameras. Much like the AI Guardman, they help authorities analyze live video streams without any need for human intervention. The system notifies law enforcement when emergency situations arise, enabling them to respond faster. The government even hopes to fine traffic offenders without the need for law enforcement to stop them. Additionally, Oyoon comes with a built-in audio system that enables cops to warn criminals that they’ve been caught in the act.
Perhaps one of the highlights of this next generation surveillance system is its facial recognition capabilities. It enables the government to track criminals who are on the run. Law enforcement can easily tail criminals around Dubai using the city-wide network of cameras.
Though terrorism and national security continue to be the top drivers of growth of the facial recognition market, recent developments have placed such technology in other various useful contexts. For example, businesses can use smart cameras to gain better insights into customer behavior. Entrepreneurs can reshape operations and ultimately improve their business based on the data they receive from the system’s analytical capabilities.
One particular industry that can greatly benefit from this technology is retail. Smart cameras can support marketers’ efforts to understand their target audiences. The systems can provide information such as age and gender of prospects to retailers.
Smart cameras are capable of detecting and classifying faces thanks to AI tech. They’re becoming increasingly powerful tools, aiding governments and even businesses in reducing crime. But these surveillance systems might not be as secure as we think.
Researchers from the University of Toronto suppose that AI-powered tech will be the downfall of smart cameras.
According to professor Parham Aarabi, an AI network built to understand how a particular surveillance system works can filter the video feed. Should the technology alter only a few specific pixels, it would be enough to make it impossible for the AI to classify the image. The researchers tested their theory in the Faces in-the-Wild Challenge. This industry-standard pool is made up of 600 different faces in various lighting conditions. The researchers introduced digital irregularities into the images to reduce the percentage of identifiable faces from 100 percent to 5 percent. However, the team concluded that there’s still a long way to go before the world sees more of this image-disrupting technology. One of the more immediate concerns: image recognition systems operate in different ways. Therefore, a single AI software can’t defeat all smart surveillance cameras.
Smart cameras certainly help businesses secure corporate assets and aid governments in preserving peace and order. The technology holds a great deal of promise. However, AI-powered tech for blocking surveillance systems is still in its infancy. Still, the current technology infrastructure in place to change the way we behave in stores and cities.
Coranet is a certified Woman-Owned Information Technology Organization that has been meeting the requirements of Enterprise and Government entities for over 30 years. Network Engineering and Technical Support is part of our foundation and we continue this proud tradition with a 360-degree approach to technology solutions and services in the areas of Audio Visual, Network & Physical Security, IT Infrastructure, and Consulting/Project Management. Coranet is ISO 9001 registered, Lean Six Sigma and WBENC certified.
Please contact us at Sales@coranet.com for more information