jiayuguan city center for disease control and prevention
shenzhen intelligent media and speech key laboratory home page research direction core technology achievements display talent recruitment about us contact us voice technology natural language processing technology embedded speech recognition technology high-performance embedded speech recognition engine with different scale vocabulary, non-specific voice recognition, natural large vocabulary with accent, continuous speech recognition the most challenging thing in speech recognition is the large vocabulary, non-specific continuous speech recognition front-end anti-noise processing background noise has a significant impact on speech recognition, its impact on speech characteristics and model parameters video retrieval
