Researchers have developed new image representations for mouse dynamics authentication that incorporate temporal information, improving performance over spatial-only methods. These representations, including XYPlot and Pairwise Distance (PWD), were evaluated using 2D-CNN and ViT architectures. The study found that location-aware PWD with temporal encoding and a 2D-CNN architecture achieved state-of-the-art results on several datasets, including Balabit and TWOS, with low Equal Error Rates (EERs). AI
IMPACT This research could lead to more robust and secure user authentication systems by leveraging advanced computer vision techniques.
RANK_REASON This is a research paper detailing a new method for mouse dynamics authentication. [lever_c_demoted from research: ic=1 ai=1.0]
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →