New extremely high-resolution 8K displays are now incorporating artificial intelligence (AI) to enhance content resolution, demonstrating both their present-day business value and their potential.
Algorithms and computing built into Samsung displays can enhance lower-resolution content — such as Full HD and 4K — by scaling them to 8K, intelligently populating all the extra screen pixels and producing stunning visuals. The AI technology from Samsung processes a video’s features while using an artificial neural network, which emulates the human brain to steadily build a stronger understanding of video characteristics.
Displays equipped with machine learning-based AI upscaling analyze and adapt content in real-time rather than stretching it to fill an 8K display, which has been the standard for years. AI upscaling populates the additional pixels and improve the display resolution. The results: enhanced, sharpened visuals that are incredibly crisp and rich, and make the case for 8K, even if the original creative was not produced for that resolution.
Showing the future now
When 4K displays first started coming on both the consumer and commercial display markets, a common concern was that while 4K material looked amazing, there was very little of it being produced. Why have a 4K screen when there wasn’t much 4K content to show?
Your ultimate video wall guide
Learn how to build the optimal video wall for your business. Download Now
Fast forward to today, and 4K video is now common.
8K is in the same situation — with a lot of excitement about its possibilities for premium visual applications. But like 4K in its early days, relatively little content is currently being produced at that super high resolution.
Explaining the basics
Whether LCD, QLED or direct view LED, commercial displays have a defined number of horizontal and vertical pixels. With flat-panel displays, resolution is a fixed number, like Full HD or Ultra HD. LED resolution may also be defined by a set size, such as the series of fixed sizes for Samsung’s The Wall. Or, resolution may be tied to the aggregate number of pixels resulting from LED “cabinets” stacked and tiled to create a video wall.
Often, video walls are a flagship retail store’s key visual feature. If a video wall needs 10 LED cabinets to fill the available wall space, and each cabinet has 300 LED light pixels horizontally, then the horizontal resolution is 3,000 pixels.
A 4K display is defined as 3,840 by 2,160 pixels, while an 8K doubles that to 7,680 by 4,320 pixels. If a 4K video was shown on an 8K display without being processed and scaled up, the video would sit in the middle of the screen, surrounded by black, only populating half of the available pixels.
When a video is upscaled or upconverted, it’s processed in such a way that matching material is created to fill in the gaps and produce a video that fills the whole screen. The onboard computing and software look at adjacent pixels within a video frame and use machine learning to determine how the new pixels should look. The content gets sharper, better-defined edges and appears more detailed.
Samsung implemented its AI 8K upscaling technology in 2018 as a result of R&D from its in-house Picture Quality Solution Lab. The goal? Using deep learning algorithms to ensure that displays are always optimizing video quality.
“Deep learning enables more precise and efficient image quality improvements than can be achieved by humans alone,” says lab team member Hyungjun Lim.
“Previous machine learning technology brought about enhanced sharpness to display picture quality, but now our technology can also offer more delicate texturing,” adds Hyunseung Lee, a lab team member who developed enhanced edge sharpness for the AI Quantum Processor 8K built into displays. “Images with complicated textures, like mountain or grass landscapes, can now be upscaled into 8K quality more naturally.”
Matching content to the available resolution is just one part of the on-screen benefits powered by AI/machine learning. The technology is also being used to reduce video background noise, repair any distortion caused by compression (used to reduce file sizes) and categorize content by video quality.
AI upscaling also creates more detailed, natural-looking video by enhancing features such as image depth, light, color and line differentiation. The results bring on-screen visuals that much closer to real life.
Native 8K consumer video content is still a largely unpursued application due to the computing overhead and infrastructure demands. 8K video comes in large files, and that affects everything from storage and processing power to bandwidth. While some consumers now have 4K TVs, a lot of the cable and streaming content they’re watching is just in converted Full HD.
4K resolution addresses the everyday needs of many commercial display applications, like retail digital signage and menu displays, but 8K screens can deliver the granular detail you want for display projects in applications such as medical imaging and research, energy exploration, museums and archives.
Imagine medical imaging on a lobby video wall at a cancer institute, showing the super-fine details that are normally only visible under a microscope. Or, imagine ancient manuscripts in an online exhibit, digitized in such fine detail that the viewer can even see minute characteristics of the paper and the ink.
Native 8K presents the future opportunity to show lifelike visuals — at life size — on LED screens. AI bridges that gap to business display projects today.