Adrian Kretz

Papers

Total Citations

H-Index

About

Adrian Kretz is a researcher focused on advancing computer vision, particularly in the domain of depth estimation and semantic scene understanding. His most notable contribution, the "SDNet: Semantically Guided Depth Estimation Network" (2019), introduces a novel architecture that leverages semantic information to improve monocular depth prediction. By integrating high-level scene understanding into the depth estimation pipeline, Kretz's work addresses a critical challenge in autonomous systems and robotics—producing more accurate and context-aware depth maps from single images. While his seminal paper has garnered 3 citations to date, its conceptual innovation has laid groundwork for subsequent studies exploring cross-modal learning in vision tasks. Kretz's approach demonstrates how semantic priors can guide geometric reasoning, offering a pathway toward more robust perception systems. His research sits at the intersection of deep learning, 3D reconstruction, and scene parsing, with potential applications in self-driving cars, augmented reality, and assistive technologies. Though early in his career, Kretz's work exemplifies a thoughtful integration of semantic and geometric cues, marking him as a promising voice in the evolving landscape of visual intelligence.