New GBU-Palm dataset advances multimodal palm attack detection research

Researchers have introduced GBU-Palm, a large-scale multimodal video dataset designed for palm presentation attack detection (PAD). This dataset includes over 21,000 videos from 105 subjects across six different acquisition environments, featuring synchronized RGB and Near-Infrared (NIR) data. The benchmark aims to facilitate the development and evaluation of robust palm PAD methods by providing leakage-controlled protocols and testing various video architectures under different environmental conditions. Initial results indicate that while RGB-NIR fusion does not consistently outperform RGB-only inputs, architectural choices significantly impact performance degradation when faced with environmental shifts. AI

IMPACT This dataset could lead to more robust biometric security systems by improving the detection of spoofed palm scans.

RANK_REASON The cluster contains an academic paper detailing a new dataset and benchmark for a specific AI research area. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• GBU-Palm
• Palm Presentation Attack Detection
• RGB color model
• RGB-NIR multispectral camera

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →