多尺度特征深度复用的显著性目标检测算法  

Deep multiplexing multi-scale features for salient object detection

作　　者：周之平[1] 樊斌 盖杉 徐温程 Zhou Zhiping;Fan Bin;Gai Shan;Xu Wencheng(School of Information Engineering,Nanchang Hangkong University,Nanchang 330063,China)

机构地区：[1]南昌航空大学信息工程学院,南昌330063

出　　处：《计算机应用研究》2022年第8期2515-2519,共5页Application Research of Computers

基　　金：国家自然科学基金资助项目(62061032)。

摘　　要：针对传统显著性目标检测方法在检测不同尺度的多个显著性目标方面的不足,提出了一种多尺度特征深度复用的显著性目标检测算法,网络模型由垂直堆叠的双向密集特征聚合模块和水平堆叠的多分辨率语义互补模块组成。首先,双向密集特征聚合模块基于ResNet骨干网络提取不同分辨率语义特征;然后,依次在top-down和bottom-up两条通路上进行自适应融合,以获取不同层次多尺度表征特征;最后,通过多分辨率语义互补模块对两个相邻层次的多尺度特征进行融合,以消除不同层次上特征之间的相互串扰来增强预测结果的一致性。在五个基准数据集上进行的实验结果表明,该方法在F_(max)、S_(m)、MAE最高能达到0.939、0.921、0.028,且检测速率可达74.6 fps,与其他对比算法相比有着更好的检测性能。In view of the shortcomings of traditional salient target detection methods in detecting multiple salient targets at different scales,this paper presented a salient object detection algorithm with deep multiplexing of multi-scale features.The network model consisted of vertically stacked bidirectional dense feature aggregation modules and horizontally stacked multi-resolution semantic complementary modules.Firstly,the bidirectional dense feature aggregation module extracted semantic features of different resolutions based on the ResNet backbone network,and then performed adaptive fusion on the top-down and bottom-up paths in turn to obtain multi-scale representation features at different levels.The multi-resolution semantic complementation module fused the multi-scale features of two adjacent levels to eliminate the mutual crosstalk between features at different levels and enhance the consistency of prediction results.The experimental results on 5 benchmark datasets show that the method can achieve the highest F_(max),S_(m),MAE of 0.939,0.921,0.028,and the detection rate can reach 74.6 fps,which has better detection performance compared with other comparison algorithms.

关 键 词：显著性目标检测 多尺度特征 双向密集特征聚合 多分辨率语义 深度学习 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]