Andrea Cohen

Papers

Total Citations

H-Index

About

Andrea Cohen’s research lies at the intersection of computer vision and robotics, with a focus on visual localization and long-term mapping for autonomous systems. Her most impactful contribution is the development of hybrid scene compression techniques that enable efficient and accurate image-to-3D model localization—a core requirement for augmented reality, drones, and self-driving cars. Her 2019 paper on this topic has garnered 65 citations, underscoring its significance in making localization practical for resource-constrained mobile devices. Cohen has also advanced the field of long-term dense mapping with her work on PlaneSDF-based change detection, which allows robots to detect environmental changes across multiple mapping sessions, ensuring conflict-free understanding of dynamic surroundings. This capability is critical for robots operating over extended periods. Through her research, Cohen addresses key challenges in deploying autonomous agents in real-world environments, balancing computational efficiency with robust performance. Her work continues to shape how machines perceive and navigate complex, changing spaces, making her a notable contributor to modern visual localization and mapping.