Recognizing song-based blink patterns: applications for restricted and universal access

We introduce a system for recognizing patterns of eye blinks for use in assistive technology interfaces and security systems. First, we present a blink-based interface for controlling devices. Well known songs are used as the cadence for the blinked patterns. Our system distinguishes between ten similar patterns with 99.0% accuracy. Second, we present a method for identifying individual people based on the characteristics of how they perform a specific pattern (their "blinkprint"). This technique could be used in conjunction with face recognition for security systems. We are able to distinguish between nine individuals with 82.02% accuracy based solely on how they blink the same pattern.