A Multi-Pressure Keyboard Key Sounds Audio Dataset

The acoustic properties of keyboard keystrokes are essential for advancing research in human–computer interaction, security, and accessibility, yet existing datasets often lack force variability, detailed metadata, or controlled recording conditions, limiting their applicability. To address these gaps, we introduce a dataset of keyboard keystroke sounds, systematically recorded under three distinct pressure levels (low, medium, and high) for all alphanumeric keys and the space bar, complemented by high-quality spectrograms and detailed metadata. These datasets enable the development of robust classification models for acoustic side-channel attack detection, behavioral biometrics, assistive sound-based interfaces, and hardware diagnostics while providing a reproducible benchmark for both academic research and industry applications.