A public repo for the datasets recorded as part of "A Practical Deep Learning-Based Acoustic Side Channel Attack on Keyboards".

Generate dataset for keystroke timings for exploratory and research purposes.

Generator of keystroke dynamics dataset

Bilingual Keystroke Dynamics Dataset

As cybersecurity attacks increase, the static methods for anomaly detection are no longer enough for protection. We proposed an algorithm for dynamic anomaly detection based on keystroke patterns that are unique for everyone, just like your fingerprint but digitally. Behavioral biometrics also overcome the most important limitation of physiological biometrics systems, as we can collect them without the knowledge of the user, allowing for continuous authentication. The user rhythm for 51 users with the fixed text password (.tie5Roanl) typed in 8 sessions for 50 repetitions per session. The proposed model Random Forrest achieve 94.09 % for F1 for this study. However, we tested the work by our own dataset which was collected and tested by typing the same password to check the user authentication.

Undergraduate thesis work: Dataset containing timings between keystrokes.

Simple C++ program to simulate keystrokes based on a given dataset. The given dataset used here: https://www.cs.cmu.edu/~keystroke/

Synthetic Dataset Generator for "Revisiting the Threat Space for Vision-Based Keystroke Inference Attacks" (ECCV 2020 Workshops)

TypePlus is a lightweight deep learning model for free-text keystroke authentication, evaluated across multiple loss functions, including a Filtered Mean Triplet Loss. It achieves state-of-the-art performance with a 2.86% Equal Error Rate (EER) on the Aalto University Keystroke Dataset, demonstrating its effectiveness in keystroke biometrics.

*Keylogger Monitor GUI* is a Python application featuring a graphical user interface for managing keylogger data. It allows users to view, search, and delete logged keystrokes through an intuitive Tkinter-based interface. The tool supports timestamp-based data retrieval and includes scrollable views to handle extensive datasets.

Built an EMG to keyboard decoding system that maps wrist muscle activity to QWERTY keystrokes. I compared a baseline convolutional model, CNN plus LSTM, CNN plus GRU, and a transformer encoder. The CNN plus GRU approach delivered the best accuracy with character error rate under 20 % and loss of 0.61 on the emg2qwerty dataset

Anomaly Detection for Keystroke Dynamics using the benchmark dataset: http://www.cs.cmu.edu/~keystroke/

Longitudinal Analysis for Anamoly detection in Keystroke Benchmark Dataset

A comprehensive analysis on How keystroke dynamics change over time time using Keystroke Dynamics Benchmark Dataset

Keystroke and Mouse Dataset collected for the study of Impersonation

creating and evaluating various deep learning architectures to predict keystrokes from electromyography signals in Meta's emg2qwerty dataset.

A behavioral biometrics authentication system that identifies users by their unique typing patterns using Random Forest ML — supports live keystroke enrollment and IKDD dataset training.

This is a KeyStroke Dynamics dataset with a group of 51 Individuals who have access to a passcode for a specific system. There are 34 columns in the dataset each describing the typing interval to type in the passcode: .tie5Roanl. Using the Known dataset we are going to predict the users in unknown and the questioned dataset.

predict overall writing quality. Does typing behavior affect the outcome of an essay? develop a model trained on a large dataset of keystroke logs that have captured writing process features.

APPLIED KEYSTROKES Datasets on CAPSNET

Keystroke Acoustic Dataset (KAD)

buat koleksi dataset keystroke

A analysis on the dataset for keystroke dynamics

Keystroke coding dataset for two web programming problems

GMM Model implementation for three mobile keystroke datasets: Coakley, Antal and Teh.

Analysis of the dataset "Observations on Typing from 136 Million Keystrokes"

Multiclass keystroke dynamics authentication (user identification) using the DS Strong Password dataset.

Phrase-based keystroke behavioral authentication using dot product similarity, benchmarked against Manhattan, Mahalanobis, and SVM methods on the CMU Keystroke Dynamics Dataset.

Keystroke Anomaly detector on CMU (Carnegie Mellon University) dataset using Euclidean Distance in Python

Track Keystroke timings and correct any errors to create a dataset for model training