User usage mode of GYD332X series speech recognition chip

Users can set two different user usage modes through programming: “trigger recognition mode” and “loop recognition mode”:

Trigger identification mode

After receiving an external trigger (such as the user pressing a button), the system’s main control MCU starts a timing recognition process (such as 5 seconds) of the GYD3320 chip, requiring the user to speak the voice to be recognized during this timing process. Keywords. After this process, the user needs to trigger it again to start a recognition process again.

Loop recognition pattern

The system’s main control MCU repeatedly starts the identification process. If no one speaks and there is no recognition result, a recognition process will be started after the timing of each recognition process expires; if there is a recognition result, a recognition process will be started again after corresponding processing based on the recognition (such as playing a certain voice as an answer) .

Generally speaking, trigger recognition is suitable for occasions where recognition accuracy is required. After being triggered by the outside world, the product can play a prompt sound or other methods to prompt the user to say the content to be identified in the next few seconds. This guides the user to only say the content to be identified within the specified time, thus ensuring comparison. High recognition rate.

Loop recognition is more suitable for occasions where voice monitoring is required all the time, or where there are no buttons or other devices to control the start of recognition. In this state, the recognition accuracy will decrease to a certain extent. During the loop recognition process, other voices of the user or other sounds from the outside may be misrecognized by the recognition engine and give wrong results, which requires the control logic of the product. All are handled accordingly.

These two methods are reflected in the Demo provided by the evaluation board, and combined with the feedback from users using the evaluation board, the speech recognition chip GYD3320 developed by Gyuan Electronics provides an in-depth analysis and summary of the two modes. And provides practical solutions in product design.