Andrew Rabinovich

Papers

Total Citations

H-Index

About

Andrew Rabinovich is a leading researcher in computer vision and deep learning, with a focus on 3D scene understanding and geometric reasoning from monocular imagery. His major contributions include pioneering end-to-end deep learning approaches for 3D object detection and indoor spatial mapping. In his highly cited work "Deep Cuboid Detection: Beyond 2D Bounding Boxes" (2016, 18 citations), Rabinovich introduced the first deep learning framework to directly localize 3D cuboids from single RGB images, moving beyond traditional methods reliant on low-level geometric cues. This work laid the foundation for robust object-level 3D understanding in cluttered scenes. He further advanced indoor perception with "DeepPerimeter: Indoor Boundary Estimation from Posed Monocular Sequences" (2019, 9 citations), which developed a novel pipeline combining deep depth estimation and wall segmentation to infer complete floor plans from simple video sequences. Rabinovich's research bridges the gap between 2D vision and 3D spatial intelligence, enabling practical applications in robotics, augmented reality, and indoor navigation. His work is distinguished by its practical, data-driven approach to solving fundamental geometric problems, making him a key figure in the evolution of modern computer vision.