Qingmao Wei
Papers
2
Total Citations
27
H-Index
2
About
Qingmao Wei is an emerging researcher specializing in computer vision and efficient deep learning, with a particular focus on real-time visual object tracking. His work addresses one of the most pressing challenges in modern robotics and edge computing: bridging the gap between high-performing transformer-based models and the computational constraints of real-world deployment environments. Wei's most notable contribution is LiteTrack, a innovative framework that employs layer pruning combined with asynchronous feature extraction to create lightweight yet highly capable visual trackers. This work directly tackles the latency bottleneck that typically accompanies performance gains in transformer architectures, making sophisticated tracking accessible for edge devices and real-time robotics applications. The research has garnered 27 citations across its publications, reflecting meaningful recognition within the computer vision community despite its recent introduction. What makes Wei's work particularly compelling is its practical orientation — rather than pursuing benchmark performance in isolation, he prioritizes deployable solutions that function under real hardware constraints. As transformer-based tracking continues to dominate the field, Wei's contributions toward efficiency-preserving compression techniques position him as a valuable voice in the ongoing conversation about making powerful AI systems genuinely accessible beyond high-performance computing environments.
Research Focus
Key Achievements
Top Papers
- 1
- 2