Ameba RTL8722 TensorFlow Lite Voice Recognition Example
This video demonstrates how Ameba RTL8722 Mini supports TensorFlow Lite.
Basically, a voice recognition neural network is trained with TensorFlow Lite to respond to certain keywords, like "Yes" or "No".
Then this neural network model is deployed to an embedded board like Ameba RTL8722-mini, where it contains an on-board microphone that can listen for the keyword in real-time and trigger different pins accordingly.
Realtek IoT/WiFi MCU Solution Official Website
TensorFlow Lite Simple audio recognition
AmebaD Arduino Github