Ya Jing

Papers

Total Citations

H-Index

About

Ya Jing is an emerging researcher at the forefront of robot learning and embodied AI, with a particular focus on leveraging foundation models to advance robotic manipulation capabilities. Her work sits at a compelling intersection of vision-language understanding, generative pre-training, and imitation learning, exploring how powerful pre-trained models can be adapted to enable more capable and generalizable robotic systems. Her most recognized contribution, "Vision-Language Foundation Models as Effective Robot Imitators" (2023, 19 citations), demonstrates how existing vision-language models can be efficiently fine-tuned to serve as robust robot imitators — a deceptively simple yet impactful insight that bridges the gap between large-scale multimodal learning and practical robotics applications. Complementing this, her work on video generative pre-training for visual robot manipulation explores how representations learned from large-scale video data can meaningfully enhance a robot's ability to understand and interact with the physical world. Together, these contributions reflect a clear and timely research vision: unlocking the potential of internet-scale pre-trained models to solve longstanding challenges in robotic learning. For students entering the fields of embodied AI or robot learning, Jing's work offers an excellent entry point into this rapidly evolving research frontier.