Circumventing Video Identification Using Augmented Reality

Explore a step-by-step tutorial on circumventing video identification using augmented reality in this 31-minute conference talk from the 35th Chaos Communication Congress (35C3). Learn how to augment video streams with computer-generated official ID cards, complete with visible watermarks, during video chat identification processes. Discover techniques for detecting credit cards printed with markers using the Aruco library in OpenCV, rendering 3D representations of ID cards with arbitrary information using OpenGL, and implementing skin detection to create masks for realistic compositing. Gain insights into the process of combining input images, card positioning, fake card renderings, and skin masks to produce convincing output images that can potentially fool call center agents during video identification procedures.

35C3 - Circumventing video identification using augmented reality