A Raspberry Pi that hunts reggaeton and attacks nearby Bluetooth speakers

What it does Reggaeton Be Gone is a Raspberry Pi 3 gadget that listens to ambient audio and runs an on-device machine-learning model—trained with Edge Impulse—to classify the reggaeton genre. When the model triggers, the board sends Bluetooth packets intended to disrupt nearby speakers. The build includes an OLED display, a push button, and a Bluetooth audio receiver, all running from a 5 V / 3 A supply.

The interesting bit The project couples audio classification with a conditional RF nuisance routine, essentially automating the decision to disrupt a speaker based on musical taste. Version 3.0 introduced a strike system to reduce false positives before attacking, which implies the author discovered that misfires are socially—and perhaps legally—expensive.

Key highlights

• On-device inference via an Edge Impulse audio classification model trained specifically for reggaeton detection
• Self-contained Raspberry Pi 3 hardware with an OLED screen and push-button input
• Version 3.0 adds on-device Bluetooth scanning, a strike-based confirmation filter, and an enclosure
• Explicitly tagged as experimental, with a disclaimer that the interference may violate local regulations
• A companion project, Pocket Gone, promises a cheaper, portable, simpler alternative at roughly $15 in parts

Caveats

• The README warns the project is experimental and that sending disruptive Bluetooth packets may be restricted or illegal in some jurisdictions
• A public .eim model file or cloneable Edge Impulse project is listed as planned but not yet available, so the training artifacts may not be fully reproducible from the repository alone
• No metrics are provided for the Bluetooth disruption range, success rate, or speaker compatibility

Verdict Worth a look if you are tinkering with edge ML audio pipelines or RF experimentation on Raspberry Pi. Treat it as a conversation piece rather than a turnkey product; the README itself warns it may violate local laws and the training artifacts are not yet fully open.