Weiyu Peng

Papers

Total Citations

H-Index

About

Weiyu Peng is a leading researcher in aerial robotics and computer vision, with a focus on advancing visual object tracking for unmanned aerial vehicles (UAVs). His work addresses critical challenges in deploying deep learning models on resource-constrained aerial platforms, particularly the trade-off between global feature encoding and local detail preservation. Peng’s most cited paper, “Local Perception-Aware Transformer for Aerial Tracking” (2022, 15 citations), introduces a novel Transformer architecture that incorporates inductive bias to enhance local feature modeling—a key limitation of standard Transformer-based trackers. This contribution has improved tracking robustness in complex aerial scenarios, such as fast motion and occlusion. Beyond this work, Peng’s research spans attention mechanisms and efficient neural networks for real-time UAV applications. His achievements include developing perception-aware frameworks that bridge the gap between state-of-the-art tracking algorithms and practical deployment on drones. With growing citation impact, Peng is recognized for pushing the boundaries of aerial tracking, making his work essential for researchers in autonomous navigation and embodied AI.