光学精密工程2025，Vol.33Issue(5)：789-801,13.DOI:10.37188/OPE.20253305.0789
半监督式野生动物夜间目标端到端检测
End-to-end recognition of nighttime wildlife based on semi-supervised learning
摘要
Abstract
This study addresses the challenges of low accuracy and efficiency in the detection of wildlife at night,as well as the difficulties associated with manual comprehensive labeling.An end-to-end recognition model for nighttime wildlife based on semi-supervised learning(SAN-YOLO)was proposed and investi-gated.A feature attention mechanism and a pixel attention mechanism were integrated within the YO-LOv8 framework to enhance the adaptability and feature representation capabilities of the detector for noc-turnal images.Subsequently,a semi-supervised training network based on a teacher-student learning para-digm was constructed,allowing the student model to learn from a substantial number of unlabeled original images by generating and appropriately assigning pseudo-labels.The efficacy of the constructed dataset was then evaluated.Experimental results demonstrate that the mean Average Precision(mAP)of SAN-YOLO reaches 69.7%with only 5%annotated data,surpassing the 59.6%mAP achieved with full su-pervision in its conventional detector and exceeding the baseline model's performance of 57.1%.Conse-quently,the proposed detection method exhibits robust performance with a limited number of labeled data-sets for nocturnal animals and validates the effectiveness of attention mechanisms in the domain of night-time object detection.关键词
目标检测/半监督学习/红外夜视/野生动物保护/师生模型/注意力机制Key words
object detection/semi-supervised learning/infrared nightvision/wildlife conservation/teacher-student model/attention mechanism分类
海洋学引用本文复制引用
卢汉,崔博伦,万华洋,张国峰,沈晨,王驰..半监督式野生动物夜间目标端到端检测[J].光学精密工程,2025,33(5):789-801,13.基金项目
北京市航空智能遥感装备工程技术研究中心开放基金课题(No.AIRSE20233) （No.AIRSE20233）
国家重点研发计划资助项目(No.2023YFF0722902) （No.2023YFF0722902）
国家自然科学基金资助项目(No.62175144) （No.62175144）
