Vision-Language Models for Fraud Detection

In this 23-minute conference talk from DSC EUROPE 24 in Belgrade, Renat Khizbullin addresses the increasing sophistication of fraud attacks on identification documents and the challenges of maintaining separate detection systems. Discover how vision-language models (VLMs) can revolutionize fraud detection by consolidating multiple tasks into a unified system. Learn about the key challenges in document fraud detection and how VLMs provide a scalable solution that streamlines processes while enhancing effectiveness. The presentation, delivered on November 20th, offers valuable insights for professionals working in security, artificial intelligence, and document verification fields.