Ivan Krasin

Papers

Total Citations

100

H-Index

About

Ivan Krasin has made transformative contributions to robotic manipulation and computer vision, with his work fundamentally reshaping how machines interact with the physical world. His research centers on developing neural network architectures that enable robots to understand and rearrange visual scenes through spatial reasoning. Krasin’s most influential contribution is the Transporter Network (2020, 100+ citations), a groundbreaking model that reformulates robotic manipulation as a sequence of learned spatial displacements. Unlike traditional approaches requiring explicit object models, this architecture rearranges deep visual features to infer pick-and-place actions, allowing robots to generalize across diverse objects and configurations. The work’s elegance lies in its simplicity—by treating manipulation as visual rearrangement rather than complex physics simulation, Krasin achieved robust performance with minimal training data. This paradigm-shifting approach has become foundational in robotics research, enabling applications from warehouse automation to assistive technologies. Beyond this flagship work, Krasin’s broader portfolio spans visual representation learning and embodied AI, consistently pushing toward systems that perceive and act in unstructured environments. His research exemplifies how thoughtful architectural design can bridge perception and action, making robots more adaptable and capable in real-world settings.