Marcel Geppert

Papers

Total Citations

H-Index

About

Marcel Geppert is a researcher at the forefront of privacy-preserving computer vision, specializing in structure-from-motion (SfM) and visual localization. His work addresses a critical tension in modern mixed reality and robotics: the need for cloud-based processing versus the imperative to protect sensitive visual data. Geppert’s major contribution lies in developing methods that enable accurate 3D mapping and camera pose estimation without exposing raw images or identifiable scene content. His seminal 2020 paper, “Privacy Preserving Structure-from-Motion,” has garnered 36 citations, laying the groundwork for secure cloud-based SfM. He advanced this line of inquiry in 2021 with “Privacy Preserving Localization and Mapping from Uncalibrated Cameras,” which tackles a fundamental limitation—the requirement for calibrated cameras—by leveraging privacy-preserving line features. Though recent, this work signals a significant step toward practical, privacy-compliant systems. Geppert’s research is pivotal for enabling trustworthy augmented reality and autonomous navigation, where user privacy cannot be an afterthought. His achievements are shaping a future where visual intelligence and data protection coexist seamlessly.