Deep fakes Detection Using Human Eye Blinking Pattern: Deep Vision

Abstract

Abstract: Deepfake technology is developing rapidly and has the ability to create fake videos that pose a threat to people. Modern depth detection methods often rely on facial disparity, but as technology advances these methods become obsolete. In this study, we propose a new method for deepfake detection based on observation of human eye blink patterns. Human eye blinking is a natural and involuntary action that is difficult to accurately replicate in deepfake video. In our study, we used the unique characteristics of individual blink patterns, which are influenced by many factors such as genetics, muscle tone and unconscious reflexes. We use computer vision and machine learning techniques to extract and identify these patterns from video clips. Our preliminary tests show good results in detecting deepfakes with high accuracy. We are focused on continuing to support the fight against the spread of fraud by focusing on a part of human behaviour that is difficult to replicate. This approach has the potential to improve existing tools for in-depth discovery and increase the overall security and reliability of multimedia content in the digital age. This research opens new avenues for the development of more robust, reliable and flexible deep recognition technologies. This represents a significant step forward in the ongoing fight against malicious misuse of electronic devices.