BLINGs
Immersive Content
- Year of Production
- 2025
- Detailed Genre
- Other
Video
Summary
AI solution that reconstructs and converts impaired speech to normal speech (Healthtech, Medical, Insurance, Rehabilitation, Education)
o Solution Excellence
- Industry-leading recognition accuracy
Excellent accuracy of 15~18% WER (Word Error Rate) for impaired voice recognition
Compared to global competitors such as Voiceitt (WER18-22%) and Google Project
Euphonia (WER 16-18%)
Improves communication efficiency and satisfaction for speech impaired users with
the lowest error rate.
(English WER is below 15% and Korean CER is below 20%)
- Speaker-specific characteristics preservation technology
Voxsory preserves the speaker's unique timbre, intonation, and emotional
characteristics in the process of converting impaired speech to normal speech,
emotional characteristics as much as possible. It doesn't just convert to a
standardized voice,
selective reconstruction that only improves intelligibility while preserving the user's
personality and voice identity.
technology that only enhances intelligibility while preserving the user's personality
and voice identity.
o Key features of the solution
- Support communication by converting impaired speech to normal speech
- Provides visualization of the improvement of impaired speech as a pathological
indicator
- Pseudonymization and privacy protection through modulated voice processing of
original voice
o Applications
Rehabilitation support in medical institutions, assistant to rehabilitation therapists,
healthcare insurance products of insurance companies, education,
Speech clarification for voice-enabled IOT devices
o Solution Excellence
- Industry-leading recognition accuracy
Excellent accuracy of 15~18% WER (Word Error Rate) for impaired voice recognition
Compared to global competitors such as Voiceitt (WER18-22%) and Google Project
Euphonia (WER 16-18%)
Improves communication efficiency and satisfaction for speech impaired users with
the lowest error rate.
(English WER is below 15% and Korean CER is below 20%)
- Speaker-specific characteristics preservation technology
Voxsory preserves the speaker's unique timbre, intonation, and emotional
characteristics in the process of converting impaired speech to normal speech,
emotional characteristics as much as possible. It doesn't just convert to a
standardized voice,
selective reconstruction that only improves intelligibility while preserving the user's
personality and voice identity.
technology that only enhances intelligibility while preserving the user's personality
and voice identity.
o Key features of the solution
- Support communication by converting impaired speech to normal speech
- Provides visualization of the improvement of impaired speech as a pathological
indicator
- Pseudonymization and privacy protection through modulated voice processing of
original voice
o Applications
Rehabilitation support in medical institutions, assistant to rehabilitation therapists,
healthcare insurance products of insurance companies, education,
Speech clarification for voice-enabled IOT devices
Company Profile
Broadcasting Music Startup Immersive Content
Braindeck
Braindeck is a company that create the finest AI Voice DNA. We are contributing to the expansion of IP profits of media & entertainment companies through conversation and singing models based on multilingual voice generation for virtual humans and characters, as well as conversation and singing through voice restoration of the deceased. Currently, we are focusing on developing intelligent conversation technology with persona and advanced deep voice detection technology to protect voice copyright, and are pioneering the global market through alliances with overseas partners such as the United States and Japan.
