The Korean language is so natural. I get goosebumps because it sounds like it was sung by the real Bruno Mars.”
Recently, a YouTube video of American pop singer Bruno Mars singing ‘High Mouth Boy’ by Korean idol group New Jeans has become a hot topic. The video exceeded 1 million views in about two weeks after being posted. The twist is that
the main character who sang the song was an artificial intelligence ( AI ), not a real singer. This was a video created by a user using AI voice technology. Netizens who listened to the song continued to admire the level of AI technology development, saying, “Korean is not foreign.” In this way, ‘AI
voice technology’ , which learns the voice of a real person and creates songs or audio books, is developing. Beyond voice recognition AI assistants such as Apple’s ‘Siri’ or Amazon’s ‘Alexa’, it has reached the level of synthesizing voices using voice data of singers, actors, and even ordinary people. The basic principles of voice AI technology are text-to-speech ( TTS ) and speech-to-text ( STT ). It literally means a technology that converts text into voice and voice into text. AI
With the development of technology, it has become possible to synthesize voice by learning the speech style of the original voice after analyzing voice data in units of words and sentences.
In Korea, startups such as Neosapiens and Humelo have jumped into the development of AI voice technology.
Founded in 2017, Neosapiens operates a ‘Typecast’ service that produces audio and video content using AI voice actors. When a user writes a sentence and selects a voice actor according to gender, age, content atmosphere, etc., a voice file that looks like a professional voice actor can be created.
Established in 2018, Humel also developed such an AI voice synthesis service ‘Prosody’.
In 2020, Naver launched ‘Clova Dubbing’, which allows dubbing voices to be applied to content with various AI voices. We are also providing a ‘Voice Maker’ service that allows you to create an AI voice immediately by recording a few sentences with your smartphone . (Provided by Naver)
As Naver and KT also introduced voice synthesis services, the era has come where even ordinary people, not AI experts, can easily create AI voices.
In 2020, Naver카지노사이트 launched ‘Clova Dubbing’, which allows dubbing voices to be applied to content with various AI voices. Clover dubbing has surpassed 1 million subscribers in just over two years since its launch. Naver also introduced a ‘Voice Maker’ service that can create an AI voice immediately by recording a few sentences with a smartphone.
Naver also used Clova’s voice synthesis technology to create the voice of its virtual human ‘Rosie’. Rosie’s voice was embodied as voice actor and demonstrator’s voice recording data. Naver produced voices through its own ‘ NES ‘ technology.
KT also launched similar ‘AI Voice Studio’ and ‘My AI Voice’ services last year. By recording 30 example sentences, you can create an AI voice that resembles your real voice.
Current AIVoice technology is used in fields such as dubbing and audio books. It is a method of producing content mainly with voice data of actors, singers, and voice actors.
Furthermore, an era in which individual voices can be applied to services is coming soon. In fact, overseas, an influencer recently synthesized his voice and offered a paid AI voice chatbot service, which became a hot topic.
American influencer Karin Margery has launched an AI voice chatbot called ‘Carin AI’ that replicates her voice and personality. It recorded sales of over $100,000 (approximately 134 million won) in the first week of release. Her movie ‘Her Her’ ( Her ), in which her human falls in love with her AI, has even been evaluated as a reality.