A CAPTCHA System with Face Landmark Multiplication Operation

Modern web and mobile applications face significant security threats from malicious automated bots, large-scale unauthorized data extraction, and server overload. Traditional CAPTCHA systems have been widely implemented to mitigate these threats; however, they often lead to an impaired user experience due to repetitive problem-solving and lengthy processing times. To overcome these limitations, this paper proposes a CAPTCHA solution that integrates Generative Artificial Intelligence (Gen AI) with object detection models and simple arithmetic operations, seamlessly combining them with web development techniques. This approach allows humans to quickly and effortlessly solve challenges using image recognition and basic multiplication, while making it difficult for automated programs to succeed―thus ensuring both security and usability. This dynamic CAPTCHA system enhances security by generating diverse and complex challenges that are difficult for bots to bypass, using Generative AI and object detection models. User evaluations show that the system provides an intuitive and efficient experience, making it a promising alternative to traditional CAPTCHA systems. It strikes a balance between security, accessibility, and user experience, offering strong defense against automated threats.