Naver Cloud announced on the 2nd that it has launched a 'Real-time Streaming' feature for its enterprise AI-based speech-to-text service, 'CLOVA Speech,' which extracts the speaker's voice from live broadcasts and generates subtitles immediately.
'Real-time Streaming' is a technology that instantly converts the speaker's words into text during live streaming videos such as live broadcasts. It supports three languages: Korean, English, and Japanese, and forms text at the word unit level of the speech.
By utilizing the 'Real-time Streaming' technology, subtitles can be delivered in real time without separate typing work. In customer centers, call contents can be immediately converted into text for monitoring, enabling faster customer response.
Naver Cloud lowered the service fees in conjunction with the launch of the new 'Real-time Streaming' feature. The cost for speech recognition and speaker recognition was reduced by 40% compared to before. Previously offered as a single pricing plan, the functions are now divided into speech recognition, speaker recognition, and event detection (recognition of applause, music, cheers, etc.), with fees segmented by function. A feature that evaluates the accuracy of English pronunciation was also added as an optional choice.
Kim Seong-hoon, AI Product Planning Manager at Naver Cloud, said, "We expect the real-time streaming feature to be highly utilized in industries requiring live broadcasts, such as broadcasters, live commerce companies, and YouTubers. We will continue to advance AI-based CLOVA services to support corporate business growth."
© The Asia Business Daily(www.asiae.co.kr). All rights reserved.
