TensorFlow Micro Speech demo for M5Stack FIRE

M5Stack port of micro-speech demo. Say "yes" or "no" to your M5Stack. M5stack would change its facial expression according to your words.

M5Stack is an ESP32-based module with TFT LCD and analog MEMS microphone.
This port uses Arduino-based M5Stack library on PlatformIO. I have tested this on my M5Stack FIRE, but M5GO would also work.

Most of the source files on this repository are derived from TensorFlow repository using
make -f tensorflow/lite/micro/tools/make/Makefile TARGET=esp generate_micro_speech_esp_project
excepting audio_provider.cc, command_responder.cc, command_responder.h and main.cpp.

M5Stack-Avatar library is used (but not included) for drawing a face on the LCD.

Since M5Stack's internal microphone is always picking up noise from the internal speaker, you should speak LOUDLY to your M5Stack. I'm sorry it may be annoying for people near you...

Visit my blog entry for details and demonstration video. (Sorry the article is in Japanese)