计算机科学与探索2019，Vol.13Issue(6)：1049-1061,13.
多尺度卷积特征融合的SSD目标检测算法
SSD Object Detection Algorithm with Multi-Scale Convolution Feature Fusion
摘要
Abstract
In this paper, it is proposed that the accuracy of small and medium objects detection of SSD (single shot multibox detector) can be improved by incorporating a modified multi-scale convolution feature fusion method. In order to enhance small objects detection, the region magnification extraction on the low-layer feature maps is adopted. Then, features are extracted from the high-layer feature maps to make the detection of medium objects better. These features are finally fused by the multi-scale convolution detection method as in the original SSD architecture. Moreover, the parameters of the present model are obtained via the parameter retraining. On the MS COCO test set, the present model shows that the mAP (mean average precision) of medium and small objects decetion is 75.1% and 40.5% respectively, which is nearly 16.3% and 23.1% higher than the original SSD.关键词
单次多框目标检测器(SSD)模型/多尺度特征融合/目标检测/深度学习Key words
single shot multibox detector (SSD)/multi-scale feature fusion/object detection/deep-learning分类
信息技术与安全科学引用本文复制引用
陈幻杰,王琦琦,杨国威,韩佳林,尹成娟,陈隽,王以忠..多尺度卷积特征融合的SSD目标检测算法[J].计算机科学与探索,2019,13(6):1049-1061,13.基金项目
The National Natural Science Foundation of China under Grant No. 61772176 (国家自然科学基金) （国家自然科学基金）
the Scientific and Technological Project of Henan Province under Grant Nos. 182102210078, 182102210362 (河南省科技攻关项目) （河南省科技攻关项目）
the Plan for Scientific Innova-tion Talent of Henan Province under Grant No. 184100510003 (河南省科技创新人才项目) （河南省科技创新人才项目）
the Scientific and Technological Project of Xinxiang City under Grant No. CXGG17002 (新乡市科技攻关计划项目). （新乡市科技攻关计划项目）
