Leonid Antsfeld

Papers

Total Citations

H-Index

About

Leonid Antsfeld is a researcher specializing in autonomous navigation, deep learning, and spatial representation for robotic systems. His work sits at the intersection of robotics and machine learning, with a particular focus on enabling agents to navigate complex real-world environments efficiently and intelligently. Antsfeld's most notable contribution, "Learning to Navigate Efficiently and Precisely in Real Environments" (2024), addresses one of robotics' central challenges: bridging the gap between simulated and real-world navigation by developing more realistic models of agent dynamics and sensing. This work, which has garnered 5 citations since publication, pushes beyond conventional model-based control approaches to deliver more robust autonomous navigation capabilities. His 2023 paper, "Learning with a Mole," explores how navigating agents can develop compact, transferable latent spatial representations without relying on explicit scene reconstruction — a meaningful step toward more generalizable and memory-efficient navigation systems, accumulating 2 citations in its first year. Though early in establishing his citation record, Antsfeld's research tackles foundational problems in embodied AI and robot autonomy. His dual focus on practical real-environment performance and theoretically grounded representation learning positions him as a promising contributor to the rapidly advancing field of intelligent autonomous systems.