My current learning and research interests focus on efficient visual attention mechanisms in computer vision, extreme quantization of vision models, inference runtime acceleration, sparse computation, and efficient model distillation.

I am also interested in autonomous exploration and target tracking for high-speed UAVs, involving reinforcement learning, path planning, multi-sensor fusion, perception, and control.

Recently, I have been actively studying Neural Field Theory and exploring how attention mechanisms can be re-understood and redesigned from a more effective biological perspective, with the goal of investigating their potential value for visual representation learning and efficient perception models.

If you have any interesting ideas related to these directions, please feel free to reach out.