Speech Recognition

This solution provides simple command recognition and speaker-independent voice recognition, and can be trained by the user to follow the command. Once the command is recognized, the device can provide sound effects based on the audio pre-stored in the flash memory on the chip. Currently supports 9 languages ​​and multiple dialects.

In terms of audio functions, the ISD9xxx series includes a Sigma-Delta ADC with 80dB SNR performance, coupled with a programmable gain amplifier (PGA) with a gain of up to 56dB to directly connect the microphone. The audio output is provided by a differential Class D amplifier (DPWM) capable of outputting 1W of power to an 8Ω speaker.

ISD9160VI can provide simple commands for a small number of words, such as "turn on the lights", "turn off the lights", "exit", "play" and other voice recognition. Once the command (keyword) is recognized, the device can pre-store it on the chip. Provides sound effects from the audio in the flash memory.

Key functions

• Speaker Independent Identification (SIR)
• Excellent accuracy for simple instructions for a small number of words
• Currently supports nine languages ​​and multiple dialects

ISD9160 also provides speaker independent speech recognition (SD) and can be trained by the user to follow instructions. During training, commands are recorded directly into the ISD9160 via the built-in microphone input. Keywords will only respond to the person who recorded the command. This feature can provide appropriate security for any system that can be customized by the user in most applications.