检索规则说明:AND代表“并且”;OR代表“或者”;NOT代表“不包含”;(注意必须大写,运算符两边需空一格)
检 索 范 例 :范例一: (K=图书馆学 OR K=情报学) AND A=范并思 范例二:J=计算机应用与软件 AND (U=C++ OR U=Basic) NOT M=Visual
作 者:游小荣[1] 李淑芳[1,2] 邓丰 YOU Xiaorong;LI Shufang;DENG Feng(Changzhou Textile Garment Institute,Changzhou 213164,China;Changzhou Key Laboratory of Eco-Textile Technology,Changzhou 213164,China)
机构地区:[1]常州纺织服装职业技术学院,常州213164 [2]常州市生态纺织技术重点实验室,常州213164
出 处:《北京服装学院学报(自然科学版)》2023年第2期78-84,共7页Journal of Beijing Institute of Fashion Technology:Natural Science Edition
基 金:常州市科技计划项目(CE20202019);江苏省高等学校大学生实践创新训练项目(202112807013Y)。
摘 要:针对传统服装图像检索方法存在检索效率低、精度不高的问题,提出了基于ResNet和迁移学习的服装图像检索方法。在互联网上选取当下流行的服装,建立长裙、短裙、风衣、T恤4类小样本数据集。借助基于ImageNet数据集训练好的ResNet152预训练模型,对小样本数据集进行迁移学习,进一步优化预训练模型中的参数,得到新的网络模型。基于新的网络模型,提取特征向量,做归一化处理,并进行相似度计算,排序输出图像检索结果,实现服装图像检索功能。为验证方法的有效性,搭建了基于颜色、纹理、边缘、无迁移学习和有迁移学习的服装图像检索测试平台。经实验对比,这种方法不但适用于小样本情况,而且平均精度、均值等指标以及检索效率优于其他方法,实用性强。Aiming at the problems of low retrieval efficiency and low accuracy in traditional clothing image retrieval methods,the paper proposed a clothing image retrieval method based on ResNet and transfer learning.It Selected the current popular clothing on the Internet,and established a small sample data set of four categories:long skirts,short skirts,windbreakers,and T-shirts.With the help of the ResNet152 pre-training model trained on the ImageNet data set,the transfer learning is performed on the small sample data set,and the parameters in the pre-training model were further optimized to obtain a new network model.Based on the new network model,the feature vector was extracted,normalized,the similarity was calculated,and the image retrieval results were sorted and output to realize the clothing image retrieval function.In order to verify the effectiveness of the method,a clothing image retrieval test platform based on color,texture,edge,non-transfer learning and transfer learning this study built.The experimental comparison shows that this method is not only suitable for small samples,but also has better average precision and retrieval efficiency than other methods,and has strong practicability.
分 类 号:TP391.41[自动化与计算机技术—计算机应用技术]
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在链接到云南高校图书馆文献保障联盟下载...
云南高校图书馆联盟文献共享服务平台 版权所有©
您的IP:216.73.216.249