This Content Is Only For Subscribers
Computer Vision (CV) is the field of research and category of applications concerned with teaching machines to see. It involves using AI algorithms to process visual information from cameras and extract understanding and insight from it.
CV (sometimes called machine vision) has huge implications for the way artificial intelligence (AI) will change our lives. It is fundamental to the way autonomous cars navigate and is revolutionizing healthcare with its ability to analyze and interpret scans and medical images. It also has controversial applications in law enforcement and surveillance.
In 2024, we can expect CV algorithms to become increasingly powerful and pervasive, along with the emergence of new and perhaps frightening use cases. But we can also expect a continued focus on efforts to understand and mitigate the risks. Here’s my overview of what I predict will be the most significant trends:
Synthetic Data and Generative AI
Generative AI is proving to be transformative across the board, and the field of computer vision is no exception. The ways in which it is likely to impact the development of computer vision technology in 2024 include its ability to generate synthetic data, which can be used to train computer vision systems (e.g., facial recognition, object detection) more cheaply and with less risk of violating privacy. It can also be used to label training data much more quickly and efficiently than by having humans manually label data, a time-consuming and expensive process.Â
3D Computer Vision
Computer vision algorithms are becoming increasingly sophisticated at capturing and analyzing 3D images. This can be spatial, such as using multiple cameras to capture different angles. It can also be time-based, using light sensors to measure the time it takes for light to reflect off the surface of an object (e.g., LIDAR). 3D imaging provides more accurate depth and distance data and can also be used to create more accurate 3D models for use in simulations and digital twins.
Edge Computing
Processing visual data directly on the device where it is captured reduces bandwidth costs and enables faster action. Computer vision use cases ranging from autonomous vehicles to intelligent security systems benefit from this ability to process data in near real-time. Small, discreet CV models that can run on low-power devices will be a key trend in 2024.
Autonomous Vehicles
Traditional self-driving car technology focuses on input from a variety of sources, including cameras, radar, and GPS. But humans are able to drive cars almost entirely by sight, so why shouldn’t computers be able to do the same? This line of thinking is becoming more prevalent as the self-driving car (and other vehicles) move closer to becoming an everyday reality. We can be sure that by 2024, breakthroughs in computer vision will quickly find their way into prototypes and even production vehicles.
Computer Vision In Healthcare
Doctors and medical researchers are using CV to speed up the analysis of images and scans to more efficiently identify and diagnose diseases. Algorithms can be trained to tell the difference between cancerous and healthy tissue and to capture patient data to aid in record keeping. It is also being used to monitor surgical procedures – one example use case is tracking the location of surgical instruments during an operation to ensure they are not accidentally left inside the patient.
Augmented Reality
CV plays an important role in augmented reality by enabling computers to understand visual information and overlay it with digital information. A flood of new augmented reality devices – possibly including the long-awaited sets from Meta or Apple – will hit the market in 2024, putting CV-augmented tools in the hands of a larger portion of the population than ever before.
Detecting Deepfakes
As AI-generated deepfakes become more convincing, we are building a world where the lines between “real” and “computer-generated” are relentlessly blurred. This has truly worrying implications for our ability to detect propaganda and disinformation. Many believe that CV has an important role to play in mitigating this threat, thanks to its ability to analyze images and spot telltale signs that they have been algorithmically generated. As AI becomes more embedded in our lives by 2024, technological solutions like CV will be an important part of the debate.
Ethical Computer Vision
Issues of bias and fairness arise in all areas of AI but have been particularly prominent in computer vision. For example, facial recognition algorithms are often found to be less effective at identifying people with dark skin, which means there is more room for error when used for surveillance or law enforcement purposes. In 2024, I believe we will also see a growing focus on privacy-oriented computer vision in the form of technologies that can be used in public areas without violating privacy (such as automatic face blurring).
Real-Time Computer Vision
The technology to extract insights from live video so that action can be taken immediately has matured in recent years, and we can expect to see more use cases in 2024. Real-time computer vision is already being used to scan crowds for signs of potential problems such as overcrowding, analyze security footage for intruders or other threats, or monitor machinery on a factory floor to detect safety hazards. As real-time algorithms become more sophisticated, we can expect many new and valuable use cases to emerge.
Satellite Computer Vision
Satellites are becoming cheaper to launch and operate, and the images they can capture are becoming more sophisticated and insightful. By applying CV technology to images from space, it’s possible to monitor many activities on Earth, including deforestation, the spread of floods and wildfires, the spread of urban sprawl, and activities in marine ecosystems such as pollution and migration. As satellite imagery becomes more accurate and detailed and CV algorithms become more sophisticated, we will gain deeper insights that will enable more timely interventions and better use of resources.
