Aju Business Daily

 
SEOUL -- Clova, the artificial intelligence platform of South Korea's webs service giant Naver, is now capable of providing a variety of information using composite sounds based on the sweet voice of Yu In-na, a 36-year-old actress who served as a radio DJ.

Naver said in a statement Friday that its AI platform can use composite voice sounds to naturally communicate and read out digital books via AI assistant speakers. Other than other ordinary digital assistant chores, Naver added a special love consulting feature. Users can ask questions like "What should we eat on our first date?" to receive answers.

Naver applied new self-developed technologies such as "Hybrid Dnn Text-to-Speech" (HDTS) to its existing AI platform to deliver more human-like natural voices. While it takes more than 40 hours for ordinary AIs to study how a person speaks and learn to create composite voice sounds, Naver's AI requires about four hours.

According to the web service giant, its AI needs about 100 recorded sentences to create composite voice sounds mimicking a specific person.

In an effort to gain competitiveness in the AI market, Naver acquires core source technologies through independent AI research centers or partnerships with foreign startups. Naver launched Clova in May 2017 along with AI voice assistant speakers which help users receive a variety of information.