Jing Yu Koh

Carnegie Mellon University, Google (United States)

Papers

Total Citations

H-Index

About

Jing Yu Koh is a rising star in artificial intelligence, whose research bridges computer vision, natural language processing, and embodied AI. His most influential work tackles the challenge of Vision-and-Language Navigation (VLN), where agents must follow human instructions to navigate photorealistic environments. In his highly cited 2023 paper, "A New Path," Koh introduced a novel approach that leverages synthetic instructions and imitation learning to overcome the scarcity of human-annotated training data, significantly improving the scalability and generalization of VLN agents—a critical step toward real-world robots that can understand and follow human commands. Beyond navigation, Koh has made impactful contributions to 3D scene synthesis. His work on "Simple and Effective Synthesis of Indoor 3D Scenes" demonstrates how to generate high-resolution, 3D-consistent novel views from just a handful of input images, enabling immersive exploration of indoor environments from far-extrapolated viewpoints. With his papers already accumulating over 50 citations in just a few years, Koh’s research is shaping the future of embodied AI and 3D vision, establishing him as a key innovator in creating intelligent systems that perceive, understand, and interact with the physical world.