Related Experiment Video
Updated: Jun 10, 2025
05:42
An Assessment Method and Toolkit to Evaluate Keyboard Design on Smartphones
Published on: October 5, 2020
3.1K
Multi-datasets for different keyboard key sound recognition.
Karwan M Hama Rawf1, Ayub O Abdulrahman1, Hana O Kamel1
1Department of Computer Science, College of Science, University of Halabja, Halabja, Kurdistan Region, F.R. Iraq.
Data in Brief
|October 11, 2024
Summary
The new Multi-Keyboard Acoustic (MKA) Datasets offer extensive keyboard sound recordings for cybersecurity research. These datasets aid in developing better defenses against acoustic threats and keylogging detection.
Area of Science:
- Cybersecurity and Human-Computer Interaction
- Acoustic Signal Processing
Background:
- Keyboard acoustic recognition is vital for cybersecurity and HCI.
- System performance is affected by platform, typing style, and noise.
Purpose of the Study:
- Introduce the Multi-Keyboard Acoustic (MKA) Datasets.
- Provide a comprehensive resource for acoustic keyboard analysis.
- Facilitate research in keylogging detection and acoustic emanation attacks.
Main Methods:
- Collected recordings from six platforms (HP, Lenovo, MSI, Mac, Messenger, Zoom).
- Structured data includes raw recordings, segmented sounds, and derived matrices.
- Utilized Praat tool for precise data segmentation and pre-processing.
Main Results:
- MKA Datasets are among the largest and most detailed in the domain.
- Data captures typing intricacies using both hands and all ten fingers.
- Ensures high-quality, dependable data for research.
Conclusions:
- MKA Datasets significantly advance keyboard sound recognition research.
- Contribute to developing robust recognition algorithms.
- Enhance defenses against acoustic-based security threats.
