Speech Recognition

• Introduction

Language is the most natural form of communication for human beings, and human-machine interaction can be achieved through speech analysis. Facing the industry's demand for AI speech recognition, including smart homes, Internet of vehicles, and speech assistants, all of which are used to facilitate smart technology applications. With the aim of developing AI cloud services, encouraging industry application expansions, and bridging gaps between technologies, the NCHC uses speech data for machine training and learning to achieve speech recognition, and uses complex algorithms for inference, developing and analyzing speech-to-text technologies. The NCHC provides a speech recognition cloud service platform to accelerate the development of complex speech network models, and provides speech recognition API services that can be applied in mobile devices, IoT, and smart homes.

• Scientific Breakthrough
  • Application of generative adversarial network to increase the accuracy of analyses.
  • Large-scale distributed training.
  • Dynamic computing resource extension and allocation.
  • Application in Di-phone segment.
    
• Industry Applications

The speech recognition cloud service is for industries to create customized speech recognition models and perform audio analyses. Speech recognition technology can be used to achieve smart control or operation, and help industries develop innovative applications for human-machine interactions.

Contact Person：Mr. Tseng
0903049@narlabs.org.tw