Hudson AI, a media-focused voice AI company, announced on the 25th that it has updated its AI dubbing solution, "Hudson Studio." This update focuses on improving emotional expression accuracy and automating the dubbing production workflow. The previous product name, "Timber," has been changed to "Hudson Studio."
Hudson Studio is a solution that integrates voice AI technologies required for media production and dubbing, including voice separation, speaker separation, speech-to-text (STT), text-to-speech (TTS), and voice conversion (VC). It can extract dialogue from video to generate subtitles or synthesize narration and dubbing with desired voice and emotion. It supports over 80 languages and has a modular structure, allowing for selection and combination as needed.
This update improves the accuracy of AI dubbing's emotional expression, automatically reflecting the character's subtle emotions and nonverbal acting elements. User editing features have also been enhanced, allowing users to fine-tune emotional tone and receive recommendations for similar emotional tags through the "Direction Tag System."
Workflow automation features have also been enhanced. AI automatically segments audio from video, recognizes speakers, analyzes scene context and emotion to generate colloquial translations, and processes everything from voice synthesis to quality control. This enabled the production of 30 dramas and 500 hours of broadcast dubbing in approximately three months.
Hudson AI plans to develop Hudson Studios into a fully automated AI dubbing agent in the future, optimizing it for high-volume, seasonal dubbing demand for OTT, FAST channels, and global short-form drama platforms.
CEO Shin Hyun-jin said, “This update allows us to reflect the speaker’s detailed emotions while reducing production time through workflow automation,” adding, “It can simultaneously address budget, quality, and schedule issues for media platforms that require mass dubbing.”
- See more related articles
You must be logged in to post a comment.