Catherine Sodalie
Papers
1
Total Citations
7
H-Index
1
About
Catherine Sodalie is a leading researcher in computer vision and augmented reality, with a primary focus on camera relocalization—a critical challenge for enabling seamless AR experiences and autonomous robot navigation. Her most cited work, "xyzNet: Towards Machine Learning Camera Relocalization by Using a Scene Coordinate Prediction Network" (2018, 7 citations), introduced a novel deep learning approach that predicts scene coordinates directly from images, aiming to bridge the gap between real-time performance and high accuracy. This contribution addresses a long-standing bottleneck in the field, where traditional methods often sacrificed speed for precision or vice versa. Sodalie’s research has advanced the practical deployment of AR systems by proposing more robust and efficient localization pipelines. Though early in her career, her work has already garnered attention for its innovative fusion of neural networks with geometric computer vision. She continues to explore how machine learning can solve fundamental spatial understanding problems, making her a rising voice in the intersection of AI and immersive technology.
Research Focus
Key Achievements
Top Papers
- 1