• Technical Name
  • Intelligent speech enhancement and tracking technology
  • Operator
  • National Cheng Kung University
  • Booth
  • Online display only
  • Contact
  • 盧韋智
  • Email
  • jimmylu.biome@bme.ncku.edu.tw
Technical Description Improving the signal-to-noise ratio of the target speech signal is crucial to the development of human-machine communication, machine hearing, and advanced hearing aids. This innovative solution consists of the beam forming technology, computer vision, deep neural network, and asynchronous update algorithms. We designed a new microphone array and electric circuit to carry the solution which can not only strength the target speech without keywords or traditional specific training of sound character but also ensure that real-time multi-speech speech separation and output. This solution can be applied further in intelligent living, communication between human and machine, and assistive technology.
Scientific Breakthrough This innovative solution integrated computer vision, beamforming technology, deep neural network, and asynchronous update algorithms into a well-designed microphone array and circuit which can not only real-time enhance and track the targetspeech but also ensure th speech procession and output instantly.
Industrial Applicability Intellige speech enhancement and tracking technology can not only real-time enhance and track the target speech but also ensure the speech procession and output instantly.It could be applied to development of hearing aids and signal processing of cochlear implant. According to the Global data, the market is expected to reach US$ 8.6 billion dollars in 2025 Annual growth rate of 6%.
The technology is also applied to machine hearing and combine speech recognition and natural language processing. It is well developed in AI robot housekeeper, smart conference system, smart speaker, interactive entertainment, smart home security system and etc.
本網站使用您的Cookie於優化網站。繼續瀏覽網站即表示您同意本公司隱私權政策,您可至隱私權政策了解詳細資訊。