Sunday, January 29, 2023
HomeElectronicsDetect vandalism utilizing audio classification on the Nano 33 BLE Sense

Detect vandalism utilizing audio classification on the Nano 33 BLE Sense


Detect vandalism utilizing audio classification on the Nano 33 BLE Sense

Arduino StaffDecember 1st, 2022

Having one thing damaged into and/or destroyed is an act that most individuals hope to keep away from altogether or a minimum of catch the perpetrator within the act when it does happen. And as Nekhil R. notes in his challenge write-up, conventional deterrence/detection strategies usually fail, which means {that a} newer sort of resolution was mandatory.

In contrast to different glass breaking sensors, Nekhil’s challenge depends on a single, cheap Arduino Nano 33 BLE Sense and its onboard digital microphone to document audio, classify it, after which alert a property proprietor over WiFi through an ESP8266-01 board. The dataset used to coach the machine studying mannequin got here from two sources: the Microsoft Scalable Noisy Speech Dataset for background noise, and breaking glass recorded on the system itself. Each of those had been added to an Edge Impulse challenge through the Studio and break up into two-second samples earlier than being processed by a Mel-filterbank Power (MFE) algorithm.

The ensuing mannequin, skilled utilizing 200 coaching cycles and slight noise additions, resulted in a formidable 92% accuracy, with some glass breaking samples being misclassified as mere noise. This was then exported to the Nano 33 BLE Sense as a library to be used in a sketch that frequently classifies incoming sounds and sends an e mail with the assistance of IFTTT if breaking glass is detected.

You may watch Nekhil’s demo video under and browse extra about this challenge right here on the Edge Impulse weblog.

You may comply with any responses to this entry via the RSS 2.0 feed.
You may go away a response, or trackback from your individual web site.



RELATED ARTICLES

LEAVE A REPLY

Please enter your comment!
Please enter your name here

Most Popular

Recent Comments