kitoweeknd/RFUAV
A benchmark dataset and deep learning models for Radio-Frequency-based drone detection and identification using CNN, YOLO, and Transformer architectures on RF signal data.
Velocity · 7d
+0.7
★ / day
Trend
→steady
star history
RFUAV provides a comprehensive benchmark dataset of Radio-Frequency recordings from 35 drone types for detection and identification tasks. The repository includes pre-trained deep learning models implementing a two-stage pipeline for drone signal detection and classification. It combines signal processing techniques (FFT/STFT) with modern neural architectures including CNNs, YOLO object detectors, and Transformers operating on raw IQ data and spectral representations.