Bokun He
Papers
1
Total Citations
10
H-Index
1
About
Bokun He is a researcher whose work sits at the intersection of remote sensing, deep learning, and efficient model design. His key research areas include knowledge distillation, manifold learning, and scene classification for remote sensing imagery. He is best known for his pioneering work on the "Knowledge Distillation of Grassmann Manifold Network for Remote Sensing Scene Classification" (2021, 10 citations), which addresses a critical challenge: deploying high-performance deep networks on resource-constrained devices like satellites and micro-robots. By leveraging Grassmann manifold geometry within a knowledge distillation framework, He developed a method that transfers rich representational knowledge from a large, complex network to a compact, efficient one—without sacrificing classification accuracy. This contribution is particularly impactful for real-world applications where computational power and memory are limited. His work bridges the gap between theoretical manifold learning and practical edge deployment, making deep learning more accessible for on-board satellite processing and autonomous micro-robotics. With a growing citation record, Bokun He is establishing himself as a key voice in making deep learning both powerful and portable for remote sensing.
Research Focus
Key Achievements
Top Papers
- 1