SENS-ASR: Semantic Embedding injection in Neural-transducer for Streaming Automatic Speech Recognition

Key Points

• Improvements in automatic speech recognition accuracy are noted after implementing semantic embedding methods, enhancing comprehension.
• A neural transducer model was utilized to integrate semantic embeddings, significantly impacting processing efficiency.
• Streaming capabilities were integrated, allowing real-time applications that can benefit diverse users and scenarios.
• The findings suggest potential advancements for speech processing systems in various real-time applications, highlighting the need for further exploration.