NAVER Advances Virtual Human Technology Development, Elevates Content Experiments
NAVER Advances Virtual Human Technology Development, Elevates Content Experiments
NAVER Advances Virtual Human Technology Development, Elevates Content Experiments
- NAVER unveiled ‘SORI’, its first virtual human capable of expressive and natural movements, as a special host on NAVER Shopping LIVE
- NAVER also created a matching voice for Korea’s first virtual influencer ‘ROZY’, expanding the virtual model’s reach
2022-05-06
NAVER Corp. (KRX: 035420) recently unveiled its first virtual human ‘SORI’ through the livestreaming commerce platform NAVER Shopping LIVE.
Jointly developed by NAVER and GIANTSTEP, a Korean content solution company, SORI made her debut in a new product launch of global cosmetic brand NARS on May 3. The 24-year-old, full 3D virtual human created with the Real-Time engine is capable of more expressive and natural motions compared to other virtual humans generated with computer graphics or deep fake technology.
SORI started her first career as a virtual influencer for millennials and Generation Z by introducing a few things for NARS cosmetics. As an influencer who likes to find her own favorites, SORI recommended spring-inspired shades she’s experienced, which have resonated with the viewers of the livestreaming show.
Recently, NAVER also created a matching voice for Korea’s first virtual influencer ‘ROZY’, which was conceived by SIDUS Studio X, a Korean creative content company. NAVER used its Natural End-to-end Speech Synthesis (NES) technology developed by NAVER CLOVA, the company’s AI division, to create ROZY’s AI voice. With only a short period of 40 minutes of recording, NES technology can generate synthetic voices that are lifelike compared to the speech synthesis technologies that require 40 to 100 hours. With its new voice, ROZY will be featured on a radio program, reading letters from the audience. It is the first time a virtual human will appear on the radio show.
NAVER is also providing voice synthesis solutions that can be used in content creation, including virtual human content. CLOVA Dubbing is NAVER’s text-to-speech (TTS) service, which is powered by NAVER’s speech synthesis technology. It offers 105 voices of different genders, ages and styles in various languages, such as English, Chinese and Spanish. The AI voice generated by CLOVA Dubbing is also used in TV shows, proving its natural speech generation capability.
In November 2021, NAVER also launched its CLOVA VoiceMaker beta service that can easily create AI voices. These are available for users on CLOVA Dubbing, where there are more than 20 voices available as of May 2022.
END