Samsung’s new AI codec lets users stream 8K content over crappy networks


As the resolution of content increases we need to find new ways to push it across crappy networks, and that’s where new codecs come in.


Interested in the Exponential Future? Connect, download a free E-Book, watch a keynote, or browse my blog.

It’s official, we’ve entered the era of 8K TVs. Around the world, sales of 8K displays are steadily increasing, and with TV manufacturers constantly adding more offerings to the mix, the 8K market is expected to continue to grow.


Windows are the perfect lens-less cameras for AI and VR headsets


There are, however, a few challenges that need to be addressed before viewers around the world will be able to enjoy 8K’s stunning visuals in their entirety. First, more 8K content will need to be produced, and second, network connections need to be made capable of supporting the seamless streaming of 8K movies and shows.

To address these issues, researchers from Samsung Research have developed an adaptive AI Codec known as AI ScaleNet. First introduced at SDC19 earlier this year this new codec could soon become your best friend as it works in the background to help deliver 8K content on networks that can typically only support only 4K speeds, and all without the need for additional infrastructure.


One camera to watch an entire city, meet WAMI


AI ScaleNet works by using Artificial Intelligence (AI) deep learning technology to minimise data loss during compression and enable 8K content to be streamed on networks with lower bandwidth capabilities. In a nutshell it compressed 8K content to 4K quality using an AI downscaler that’s then transmitted to the user’s TV, which then itself utilises AI to upscale the content back to 8K quality.

When asked to describe the inspiration for the technology Kwangpyo Choi, who led the project, specified a need for a new video compression and transmission technology in order to accommodate the ever-evolving media landscape, and to address related technical challenges.


Deepfake President Trump looses itself in Eminem’s raps


“Multimedia content is rapidly moving online, so figuring out how to overcome bandwidth limitations for UHD content transmission has become a major task,” said Choi. “With AI ScaleNet, users will be able to enjoy 8K-quality content, even at lower bandwidths,” he added. “In fact, during periods of high network traffic, when speeds tend to drop, viewers can still expect content to be presented at a relatively higher quality.”


Choi and Park showing off their newest codec


AI ScaleNet addresses bandwidth limitations within modern network infrastructures with what’s known as an adaptive AI Codec.

“AI Adaptive Bit Rate Streaming refers to AI technology that adapts to various bandwidths and adjusts the resolution to enable seamless streaming,” said Choi.


DeepMind's protein folding AI targets Covid-19 to help find vaccines


The difference the technology makes is especially clear when using over-the-top (OTT) services, where viewers typically notice changes in picture quality. As Choi explained, however, Samsung’s AI Codec and AI ScaleNet technology utilise AI that adapts to changes in bandwidth and optimizes quality, so users enjoy the highest-quality streaming experience possible. As the researchers described, the development process wasn’t without its difficulties.

“The most difficult part was ensuring compatibility with existing codec compression technologies,” said Choi. “How to design an easy-to-use algorithm that would be compatible with existing technologies, while delivering the intended performance at the same time was the question.”


Pentagon tests Artificial Intelligence electronic warfare systems


Choi also noted difficulties that derived from gaps in perspective between the research and development teams. There were several considerations related to complexity, design and structure which researchers took for granted, and were questioned by the developers.

“We were able to achieve this result thanks to the synergy of our team members who all worked closely to achieve the same goal, and colleagues who were focused on the technology’s commercialisation,” said Choi.

“I believe that AI ScaleNet is the first technology to overcome bandwidth limitations by addressing complexity issues and separating the server and client sides, respectively. It’s relatively difficult to commercialise AI technology today for the purpose of enhancing resolution, which is different from commercialising AI for recognition and classification.”


Google AIs learnt how to encrypt their communications to stop people spying on them


When asked how AI technology will contribute to the advancement of TV technologies in general, the researchers pointed to two possible directions.

“The first would be greater user convenience, and the second would be enhanced picture quality,” Choi began. “The first point is already being delivered with Bixby, Samsung’s digital personal assistant, to enhance people’s viewing experiences, technologies that support the best possible picture quality at minimum bandwidth, such as AI ScaleNet, should be continuously researched and developed.”

As the researchers noted, Samsung has big plans for this versatile technology.


Google's AI can rate photos on aesthetic appeal


“AI ScaleNet has shown us that by overcoming challenges, we can help usher in the future and enrich the user experience,” said Choi. “Going forward, we will continue to research and develop additional functionalities that add to users’ convenience, and through cooperation with content service providers, we hope to empower diverse services for Samsung devices.”

Related Posts

Leave a comment

Get your FREE! XPU Introduction to Exponential Thinking Course now. No registration, no catches, just awesome knowledge.GET FUTURED

Explore More!

Explore 1000's of articles about our exponential future, 1000's of pages of insights, 1000's of videos, and dig into 100's of exponential technologies. Subscribe to get your no-nonsense briefing on all the biggest stories in exponential technology and science.

Awesome! You're now subscribed.

Pin It on Pinterest

Share This