SK Telecom to showcase deep learning media upscaling solution

SK Telecom will showcase at Mobile World Congress its new media quality improvement solution that uses deep learning technology, the company has announced.

The solution, dubbed Super Nova, uses a deep learning algorithm to drastically reduce the time to process sound or videos and improve their quality, the carrier said.

It can process existing media files, as well as those filmed by camera or detected by industrial sensors.

SK Telecom said the deployment of 5G is expected to increase demand for high resolution media.

The solution was piloted by South Korean semiconductor maker SK Hynix, which used its quality assessment process in October last year. Hynix had achieved a significant reduction in the processing time of wafer images, SK Telecom said.

Quality assessment of wafers require high-resolution videos and Hynix previously filmed the same wafers multiple times and meshed the images together but Super Nova changed this, the telco said.

Super Nova has also been applied to SK Broadband's Internet Protocol TV service since July last year which allowed the latter to upscale the resolution of its broadcasts.

It has also been applied to media database of SM Entertainment, South Korea's largest entertainment firm, to increase the quality of audio. 

Last month, the company said it was collaborating with the country's capital Seoul to deploy an intelligent transportation system that uses 5G networks.

With Postech, it announced a new antenna technology that increases call quality in 28GHz spectrum in 5G.

Related Coverage

• Seoul and SK Telecom to use 5G to prevent jaywalking
• SK Telecom and POSTECH develop 5G antenna tech
• SK Telecom and Samsung to collaborate on 5G for enterprise
• SK Telecom and Sinclair to invest $33m in ATSC 3.0 joint venture
• SK Telecom, Hyundai, and Trimble to use 5G in construction
• KT and SK Telecom to end WiBro service within the year