Fig. 28.

Architecture diagram of a novel medical speech analysis system based on active perception. The system's intelligence is improved from three aspects, i.e., active hardware, active software, and human-computer interaction. Active hardware refers to the acquisition of voice information using a microphone array with active attitude adjustment ability. Active software refers to the application of reasoning and learning technology to voice recognition to continuously improve its accuracy. Finally, human-computer interaction refers to doctors and patients interacting with the voice system through voice, text, keyboard, mouse, wearable devices, etc., to realize a personalized intelligent voice system and improve the intelligence of the system.