Subject	Computer Science / Applied Machine Learning
Specific subject area	Electric Guitar Playing Technique Classification and Recognition
Data format	Raw
Type of data	.wav (audio files), .mp4 (video files), .pdf (guitar exercises), .mscz (guitar exercises), .json (train/test splits)
Data collection	A total of 549 video samples, representing nine distinct guitar techniques, are acquired over a one-month period. The recording process occurs within a home studio setup, utilizing an Android smartphone device. The setup includes the use of three unique guitars and three amplifier simulations, carefully selected to capture a diverse range of sounds. The audio WAV files are extracted from the initial MP4 video recordings using FFmpeg software.
Data source location	Institute of Informatics and Telecommunications, NCSR ‘Demokritos’, 27, Neapoleos str &, Patriarchou Grigoriou E, Ag. Paraskevi 153 41, Athens, Greece.
Data accessibility	Repository name: guitar_style_dataset Data identification number: 10.5281/zenodo.10075352 Direct URL to data: https://zenodo.org/doi/10.5281/zenodo.10075351