Real-time Speech Recognition Solution
- Latest trend learning system based on
high-quality learning big data integration of voice and text
-
- Apply a learning methodology that responds to unknown vocabulary and unstructured input according to data augmentation based on large-scale broadcast data and transcription data
- Application of deep learning post-processing technology that generates high-quality transcripts based on deep learning pre-learning giant language models
- Application of normalization technology based on a large language model for number sounds and English vocalization
- Applying robust transfer learning technology to dialect, outliers, and various background sound effects
Features
Application of world-class voice recognition technology
- Application of end-to-end (E2E) technology, the latest trend in speech recognition
- Securing performance and safety by transferring proven technology developed by ETRI
Provide real-time streaming voice recognition service
- Real-time streaming end-to-end voice recognition service available
- Expression of recognition results within 1 second after vocalization in word units
Provides a higher voice recognition rate compared to existing technologies
- Maintaining a recognition rate of about 80% or more before domain learning and providing a high recognition rate of 94% or more after domain learning with Transformer's excellent speech recognition accuracy
- Recognition rate can be strengthened in a short period of time through prior learning applied with Solugate’s patented reliability engine and automatic learning engine
Securing effectiveness with light and flexible technology
- Lightweight by applying Grammar Hybrid engine
- Can be used offline by installing on a laptop (Window) or tablet (Android)
- Supports voice recognition in various ways (real-time, batch, 8K, 16K, etc.) for more than 13 languages
Field of application of technology
Real-time voice conversion system
- Real-time subtitle service for live broadcasts or conferences
- Real-time dialogue subtitle service for the hearing impaired
- Voice recording service in specialized fields such as counseling log and medical record
Interactive Elderly Care System
- AI voice recognition based baby care system
- AI voice recognition based silver care system
- Interactive helper system for the socially vulnerable
Conversational AI Assistant
- AI speaker, smart home, smart car, smart home appliance, etc.
- AI voice recognition based learning assistance service
- AI voice recognition based kiosk system
Intelligent Contact Center
- AI voice recognition based call center system
Features
HW
- Server environment establishment, device embedded support
OS
- Linux, Windows, Android support
