模式识别与人工智能
2025年4月2日 星期三   首 页     期刊简介     编委会     投稿指南     伦理声明     联系我们                                                                English
模式识别与人工智能  2024, Vol. 37 Issue (5): 424-434    DOI: 10.16451/j.cnki.issn1003-6059.202405004
目标检测识别与对抗 最新目录| 下期目录| 过刊浏览| 高级检索 |
基于多层次融合的弱监督目标检测网络
曹环1, 陈曾平1
1.中山大学 电子与通信工程学院 深圳 518107
Multi-level Fusion Based Weakly Supervised Object Detection Network
CAO Huan1, CHEN Zengping1
1. School of Electronics and Communication Engineering, Sun Yat-Sen University, Shenzhen 518107

全文: PDF (3582 KB)   HTML (1 KB) 
输出: BibTeX | EndNote (RIS)      
摘要 由于缺少精确的边界框注释,弱监督目标检测器依赖预训练图像分类模型对候选区域进行分类.然而,预训练模型通常对具有鉴别性的区域而非完整的目标产生高响应,导致局部主导、实例丢失和非紧密框等问题.为此,文中提出基于多层次融合的弱监督目标检测网络,从增强对弱鉴别性空间特征的学习、类内样本特征丰富性和可信伪标签权重的角度提升检测性能.首先,幂池化层利用幂函数加权融合邻域内的激活值,减少弱鉴别性特征的信息损失.其次,特征混合方法随机融合候选区域的特征向量,丰富训练样本特征的多样性.最后,基于置信度的样本重加权策略融合预测值和伪标签的置信度,调节伪标签对训练的影响.在3个基准数据集上的实验表明文中网络性能较优.
服务
把本文推荐给朋友
加入我的书架
加入引用管理器
E-mail Alert
RSS
作者相关文章
曹环
陈曾平
关键词 目标检测弱监督学习多层次融合深度网络    
Abstract:Due to the lack of precise bounding box annotations, weakly supervised object detectors rely on the pretrained image classification model to classify candidate regions. However, the pretrained model often produces high responses for discriminative regions rather than complete objects, resulting in the problems of part domination, instance missing and untight boxes. To address these issues, a multi-level fusion based weakly supervised object detection network is proposed. The detection performance is improved from the perspectives of enhancing the weak discriminative spatial feature learning, enriching intra-class sample features and weighting reliable pseudo-labels. Firstly, a power function is utilized to weight and fuse the activation values within the neighborhood by the power pooling layer to reduce information loss of weak discriminative features. Secondly, the feature vectors of candidate regions are randomly fused by the feature mixing method to enrich the diversity of training sample features. Finally, the confidence of predictions and pseudo-labels is fused via the confidence-based sample re-weighting strategy to adjust the influence of pseudo-labels on training. Experiments on three benchmarks demonstrate the superiority of the proposed network.
Key wordsObject Detection    Weakly Supervised Learning    Multi-level Fusion    Deep Network   
收稿日期: 2024-02-16     
ZTFLH: TP391.4  
通讯作者: 陈曾平,博士,教授,主要研究方向为空间态势感知、软件化雷达探测、宽带成像识别.E-mail:chenzengp@mail.sysu.edu.cn.   
作者简介: 曹 环,博士研究生,主要研究方向为深度学习、计算机视觉.E-mail:caohuan_sysu@163.com.
引用本文:   
曹环, 陈曾平. 基于多层次融合的弱监督目标检测网络[J]. 模式识别与人工智能, 2024, 37(5): 424-434. CAO Huan, CHEN Zengping. Multi-level Fusion Based Weakly Supervised Object Detection Network. Pattern Recognition and Artificial Intelligence, 2024, 37(5): 424-434.
链接本文:  
http://manu46.magtech.com.cn/Jweb_prai/CN/10.16451/j.cnki.issn1003-6059.202405004      或     http://manu46.magtech.com.cn/Jweb_prai/CN/Y2024/V37/I5/424
版权所有 © 《模式识别与人工智能》编辑部
地址:安微省合肥市蜀山湖路350号 电话:0551-65591176 传真:0551-65591176 Email:bjb@iim.ac.cn
本系统由北京玛格泰克科技发展有限公司设计开发 技术支持:support@magtech.com.cn