Yuejiao Fei

Papers

Total Citations

H-Index

About

Yuejiao Fei’s research lies at the intersection of computer vision and robotic manipulation, with a core focus on enabling real-time, data-driven grasping using deep learning. Her work addresses a critical challenge in robotics: how to make autonomous grasping robust, efficient, and practical for real-world environments. Fei’s major contributions center on developing novel neural network architectures that leverage limited sensory data—such as depth-only or 2.5D images—to predict grasp quality and pose with high accuracy. Her 2019 paper, “UG-Net for Robotic Grasping using Only Depth Image,” introduced a real-time convolutional encoder-decoder network that achieves pixel-wise grasp prediction, a significant step toward simplifying perception pipelines. This work has garnered 7 citations and laid the groundwork for her subsequent studies on multimodal RGB-D fusion and height-robust grasping. In “Deep Robotic Prediction with hierarchical RGB-D Fusion” (4 citations), she proposed a novel fusion strategy that balances speed and accuracy, while her “2.5D Image-based Robotic Grasping” (2 citations) demonstrated how sampling from real sensors can enhance policy robustness across varying observation heights. Collectively, Fei’s research advances the goal of making robotic grasping more accessible and reliable, with direct implications for industrial automation and service robotics.