Analysis of fast prototyping of microcontroller-based ML software for acoustic signal classification
[ 1 ] Wydział Informatyki i Telekomunikacji, Politechnika Poznańska | [ 2 ] Instytut Automatyki i Robotyki, Wydział Automatyki, Robotyki i Elektrotechniki, Politechnika Poznańska | [ S ] student | [ P ] employee
[2.2] Automation, electronics, electrical engineering and space technologies
2023
chapter in monograph / paper
english
- TinyML
- Edge Impulse
- Nordic Thingy
- STM32
- ArduinoNano 33
EN The paper analyzes the preparation of software for acoustic signal classification with machine learning techniques for microcontrollers. The design process was tested for three types of devices: Nordic Thingy:53, SensorTile.box and Arduino Nano 33 BLE Sense Lite. The classifier training process was carried out using the Edge Impulse platform. Experimental studies were carried out for the process of classifying sound signals generated by the vacuum cleaner motor. The results of the training and the model test were presented for different configurations.
36 - 41
20