Christoph Busch

Multi-VLM fusion not only boosts face recognition accuracy but also delivers richer, more robust explanations that enhance transparency and auditability.

JPEG AI emerges as the top performer for face recognition in ultra-compressed images, revealing that significant compression can be achieved without sacrificing accuracy.

T-shirt presentation attacks can fool face recognition systems, but a simple spatial consistency check between detected faces and persons offers a surprisingly effective defense.

Vision foundation models can be efficiently adapted to dramatically improve morphing attack detection, slashing error rates by nearly 3x at critical security thresholds.