Skip to content

Latest commit

 

History

History
6 lines (4 loc) · 394 Bytes

README.md

File metadata and controls

6 lines (4 loc) · 394 Bytes

TFG

This is my undergraduate thesis about a microcontroller-based automatic baby cry detection system.

It's based on a NXP Kinetis KL25 microcontroller running a neural network inference.

The neural network is fed MFCC (Mel-frequency cepstrum coefficients) extracted from audio data captured from the board's microphone and trained against baby cry samples to detect silence vs baby cries.