Sitara Linux ALSA DSP Microphone Array Voice Recognition
Voice recognition applications requiring a superior user experience typically use a number of digital signal processing algorithms to filter, clean, and enhance microphone sampled audio. Frequently these systems will employ multiple microphones where more sophisticated algorithms can be used. These systems can leverage a Texas Instruments Digital Signal Processor to offload the complex real-time computation from the main applications processor. This application report details the hardware and software interfacing necessary to integrate a Texas Instruments Sitara ARM processor running the TI Processor Linux® SDK with the Texas Instruments C5000™ DSP evaluation module running a fixed-function four-microphone beamforming demo application as an Advanced Linux Sound Architecture (ALSA) device.