Description
Scanning with an SDR, it's hard to tell signals apart by ear or waterfall. RTL-ML is an AI radio signal classifier using RTL-SDR and machine learning with 96.9% accuracy.
It identifies FM broadcast, NOAA weather, APRS, FRS/GMRS, ISM sensors and pagers, with a full capture-train-classify pipeline on a Raspberry Pi 5.
Classification:Common signal types.
Accuracy:96.9%.
Pipeline:Capture to classify.
SBC:Raspberry Pi 5.
It identifies FM broadcast, NOAA weather, APRS, FRS/GMRS, ISM sensors and pagers, with a full capture-train-classify pipeline on a Raspberry Pi 5.
Features
Classification:Common signal types.
Accuracy:96.9%.
Pipeline:Capture to classify.
SBC:Raspberry Pi 5.
