Pengju Ma

Papers

Total Citations

H-Index

About

Pengju Ma is a computer vision researcher whose work focuses on advancing pedestrian detection systems, particularly for challenging low-light environments. His key research areas include infrared imaging, lightweight neural network architectures, and real-time object detection. Ma’s most notable contribution is the development of LE-G2F-YOLOv8, an innovative infrared pedestrian detector that integrates low-frequency enhancement techniques with a lightweight design. This work addresses critical limitations of traditional visible-light-based methods, which often fail under poor contrast and occlusion conditions common in nighttime or adverse weather scenarios. By optimizing the YOLOv8 framework with ghost modules and frequency-domain processing, Ma’s approach achieves robust detection performance while maintaining computational efficiency—a crucial balance for deployment in intelligent surveillance and autonomous driving systems. Though his 2025 paper has garnered early citations, reflecting growing interest in practical, low-light detection solutions, Ma’s research demonstrates a clear commitment to bridging the gap between academic innovation and real-world application. His work holds particular promise for advancing safety-critical technologies where reliable pedestrian detection in darkness is essential.