Computer Vision/Machine Vision is one of the most exciting applications of artificial intelligence. Algorithms that can understand moving images and videos are the key technology behind many innovations, from automation systems and self-driving vehicles to smart industrial machinery and even filters on your phone to make your Instagram photos look better.

Along with language processing (natural language processing, or “NLP”), it underpins our efforts to create machines that can understand and learn about the world around us as we do. It generally involves applications powered by deep learning—neural networks trained on thousands, millions, or billions of images until they become experts at classifying what they can “see.”

The market value of computer vision technology is predicted to reach $48 billion by the end of 2022 and is likely to be a source of continued innovation and disruption throughout the year. So, let’s take a look at some of the key trends we’ll be watching in this exciting technology:

- Data-centric computer vision

Data-centric artificial intelligence is based on the idea that the focus should be on optimizing the quality of the data used to train algorithms as well as the development of models and algorithms. Advocated by Andrew Ng - a well-known pioneer of deep learning - this emerging paradigm is relevant across AI fields but especially in the field of computer vision. Some of the first deep learning-based image recognition models were developed by Dr. Ng at Google, with the aim of training computers to recognize images of cats, and they are highly dependent on the quality of the data they are given, rather than just the quantity. Focusing on iteratively improving the quality of labeling—using techniques that automatically extract and label data—will enable computer vision technology to be applied to problems where less data is available, potentially reducing costs (in terms of money as well as computing resources) and opening up many new potential use cases.

- Computer Vision in Health and Safety

An important use case for computer vision is hazard detection and alerting when something goes wrong. Methods have been developed to allow computers to detect unsafe behavior on construction sites – such as workers not wearing hard hats or safety harnesses, as well as monitor environments where heavy machinery such as forklifts are operating in close proximity to humans, allowing them to automatically shut down if someone steps into their path. With 2.7 million people injured in workplace accidents each year, according to the US Bureau of Labor Statistics, this is an area where businesses are increasingly investing to reduce the human and financial costs of poor monitoring or inattention.

Of course, preventing the spread of viral diseases is also an important use case today, with computer vision technology increasingly being deployed to monitor compliance with social distancing requirements, as well as mask mandates. Computer vision algorithms have also been developed during the current pandemic to aid in the diagnosis of infections on chest X-rays by looking for evidence of infection and lung imaging lesions.

• Computer Vision in Retail

Shopping and retail are other areas of life where we’re sure to see computer vision technology grow in popularity in 2022. Amazon has pioneered the concept of cashierless stores with its Go grocery stores, which are equipped with cameras that can tell which items customers are taking from the shelves. More stores are set to open throughout 2022, and other retailers are set to jump on the bandwagon, including Tesco, which is opening the UK’s first checkout-free supermarket.

In addition to relieving humans of the responsibility of scanning purchases, computer vision has a number of other applications in retail, including inventory management, where cameras are used to check the stock on shelves and in the warehouse and automatically order refills when necessary. It is also used to track and understand customer movement patterns around stores to optimize merchandise positioning and, of course, in security systems to deter store vandals. Another increasingly popular use case involves allowing customers to get information about a product by scanning a barcode with their mobile phone.

In fashion retail, a particularly interesting application of computer vision is the “virtual fitting room” that allows shoppers to virtually try on items without touching them – cameras in the mirror simply superimpose images of the garment onto the mirror’s reflection and can even identify the item the customer is trying on and suggest appropriate accessories to go with it.

- Computer Vision in Autonomous and Connected Cars

Computer vision is an integral part of the connected systems in modern cars. While our first thoughts may be about the upcoming autonomous cars, it has a number of other uses in the range of “connected” cars that are currently on the road and parked in garages. Systems have been developed that use cameras to monitor facial expressions to look for warning signs that we may be tired and at risk of falling asleep at the wheel. Since this is thought to be a factor in up to 25% of serious and fatal road accidents, it’s clear that measures like this could easily save lives. The technology is already being used in commercial vehicles like delivery trucks, and by 2022, we could see it start to make its way into personal cars. Other proposed applications for computer vision in cars that could make it from the drawing board to reality include monitoring whether seatbelts are worn and even whether passengers have left their keys and phones behind when leaving taxis and rideshare vehicles. Of course, computer vision will also play a major role in self-driving cars—the current thinking is that it will be the most important element of autonomous navigation. This year, Tesla announced that its cars will rely primarily on computer vision rather than lidar and radar, which use lasers and radio waves, respectively, to build a model of the car’s environment.

- Computer Vision in Edge Computing

Edge computing describes systems where computation is performed as close to the source of the data as possible. It’s a term used in contrast to the cloud computing model, where data is collected via sensors and sent to centralized servers for storage and processing. In the field of computer vision, it’s an increasingly useful concept, as computer vision systems often perform tasks that require immediate action (think of the use cases discussed in this article under the heading of safety and autonomous driving), and simply the space and time it takes for data to be sent to the cloud is a concern!

As well as the speed gains that can be achieved, edge computing in relation to computer vision has important implications for security – an important consideration as businesses and individuals face increased scrutiny and regulation over how video data is collected and used. With advanced devices like security cameras equipped with computer vision, the data can be quickly analyzed and discarded if there is no reason to keep it, such as if no suspicious activity is detected.

Article source: https://www.forbes.com/sites/bernardmarr/2022/03/04/the-5-biggest-computer-vision-trends-in-2022/?sh=5740010f19b3