Mona Sheikh Zeinoddin
Papers
1
Total Citations
4
H-Index
1
About
Mona Sheikh Zeinoddin is a rising researcher at the intersection of computer vision and robotic-assisted surgery (RAS). Her work focuses on solving the critical challenge of accurate depth estimation for 3D reconstruction and visualization in minimally invasive procedures. She is best known for introducing **DARES**, a novel framework that adapts powerful foundation models—specifically Depth Anything Models (DAM)—to the unique constraints of surgical environments. Rather than costly full fine-tuning, her self-supervised Vector-LoRA approach efficiently tailors these models to limited surgical data, achieving robust performance without catastrophic forgetting. This breakthrough, published in 2025 and already garnering 4 citations, addresses a key bottleneck in RAS: the domain gap between natural images and the complex, texture-poor scenes inside the body. By enabling more reliable depth perception, Zeinoddin’s work directly enhances intraoperative navigation and visualization, paving the way for safer, more autonomous robotic surgery. Her innovative blending of foundation model transfer learning with surgical domain adaptation marks her as a promising voice in next-generation medical robotics.
Research Focus
Key Achievements
Top Papers
- 1