I'm building a micropython firmware for running tensorflow lite for microcontroller examples. One of the main reference examples is called micro_speech and involves reading audio from a MEMS microphone at 16Khz using DMA and I2S and then feeding it ...