Publications by authors named "Xiao-Mei Tu"
Article Synopsis
- A new wildlife target detection algorithm is introduced, based on the improved YOLOX-s network, aimed at enhancing detection accuracy in challenging rainy and nighttime conditions.
- The algorithm integrates three key components: the MobileViT-Pooling module for efficient feature extraction, the Dynamic Head module for improved task-specific detection, and the Focal-IoU module for better loss function handling.
- Experimental results show significant improvements in detection performance, with mAP scores increasing by 7.9% and 5.3%, demonstrating enhanced accuracy and reduced false detections for wildlife.
Object detector based on fully convolutional network achieves excellent performance. However, existing detection algorithms still face challenges such as low detection accuracy in dense scenes and issues with occlusion of dense targets. To address these two challenges, we propose an Global Remote Feature Modulation End-to-End (GRFME2E) detection algorithm.
View Article and Find Full Text PDFOsteoblasts are a prerequisite for osteogenesis and bone formation, and play a key role in metabolic balance, growth, development and wound repair. G protein-coupled receptor kinase interacting protein 1 (GIT1) and a series of miRNAs are known to have important effects in the growth and migration of osteoblasts, but little is known about micro RNAs (miRNAs) targeting GIT1. The present study found that miR-125a-3p has matching sites on GIT1.
View Article and Find Full Text PDF