Erkut Erdem

Papers

Total Citations

H-Index

About

Erkut Erdem is a computer vision and machine learning researcher whose work spans visual perception, synthetic data generation, and physical reasoning in robotic systems. His research addresses challenging real-world problems, including person tracking under adverse weather conditions, where he has explored the use of synthetic data as a powerful tool to overcome the scarcity of labeled training data in difficult environments — a contribution that has garnered 17 citations and highlights the growing importance of simulation-to-real transfer in computer vision. More recently, Erdem has turned his attention to robotic manipulation, specifically the problem of push effect prediction, where understanding object and scene relationships is critical for enabling robots to reason about non-prehensile actions such as pushing. His work on object and relation centric representations provides a structured framework for predicting the physical consequences of pushing interactions, contributing to the broader fields of robot learning and physical scene understanding. With publications appearing across multiple years and accumulating citations in both computer vision and robotics communities, Erdem represents an emerging voice bridging perception and embodied intelligence — making his research particularly relevant for students exploring the intersection of deep learning and autonomous systems.