Georgia Gkioxari
Papers
6
Total Citations
149
H-Index
4
About
Georgia Gkioxari is a computer vision researcher whose work centers on 3D scene understanding, object detection, and multimodal perception. She is best known for her landmark **Omni3D** project, which established a large-scale benchmark and unified model for 3D object detection from single images across diverse real-world environments — a contribution that has already garnered over 89 citations and directly addresses longstanding limitations in scalability that have hindered 3D vision research compared to its 2D counterpart. Her work spans the full pipeline of 3D understanding: from recovering object shape and spatial layout from 2D images without 3D supervision, to pioneering the fusion of vision and touch for 3D shape reconstruction, drawing inspiration from how humans naturally explore objects. More recently, Gkioxari has expanded into thermal imaging, contributing the Caltech Aerial RGB-Thermal Dataset and work on monocular depth estimation from thermal data — capabilities critical for robotics operating in low-visibility conditions. Across her portfolio, she consistently pushes toward more generalizable, data-efficient, and physically grounded perception systems, making her a significant voice in bridging 3D computer vision with real-world robotic and augmented reality applications.
Research Focus
Key Achievements
Top Papers
- 1Omni3D: A Large Benchmark and Model for 3D Object Detection in the Wild89 citations · 2023
- 23D Shape Reconstruction from Vision and Touch25 citations · 2020
- 3Learning 3D Object Shape and Layout without 3D Supervision18 citations · 2022
- 4Caltech Aerial RGB-Thermal Dataset in the Wild11 citations · 2024
- 5Omni3D: A Large Benchmark and Model for 3D Object Detection in the Wild4 citations · 2022
- 6