Automatic Image Annotation for Human Machine Interaction in Industrial IoT Flexible Manufacturing

作     者：Wang, Xiaojie Zheng, Guifeng Wu, Yu Xiong, Xuanrui Zhou, Guanghai Tolba, Amr Ning, Zhaolong 

作者机构：Chongqing University of Posts and Telecommunications School of Communications and Information Engineering Chongqing400065 China Chongqing University of Posts and Telecommunications School of Cyber Security and Information Law Chongqing400065 China Dalian University of Technology School of Software Dalian116024 China King Saud University Community College Department of Computer Science Riyadh11437 Saudi Arabia 

出 版 物：《IEEE Internet of Things Journal》 (IEEE Internet Things J.)

年 卷 期：2024年

核心收录：

学科分类：1201[管理学-管理科学与工程(可授管理学、工学学位)] 0808[工学-电气工程] 081203[工学-计算机应用技术] 08[工学] 0835[工学-软件工程] 0802[工学-机械工程] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

主　　题：Smart manufacturing 

摘      要：With the explosive growth in Industrial Internet-of-things (IIoT) devices, the volume of multimedia data in the field of flexible manufacturing has also increased significantly in recent years, especially the vast amount of unlabelled image data. Image annotation provides machines with a more natural way to interact with users, enhancing the level of intelligence in IIoT flexible manufacturing. This paper proposes a multi-feature fusion multi-kernel learning image annotation method to tackle imbalanced label distribution, image weak labeling, and varying representational abilities of features. Initially, oversampling techniques with synthetic minority class samples address the influence of minority classes, while a label enhancer extends label vectors to overcome the influence of weak labeling. Subsequently, the integration of traditional visual features with deep features based on multi-kernel learning is investigated to enhance feature representation capability. This approach combines complementary information from multiple features, establishing intrinsic connections between images and annotated keywords. Experimental evaluations are conducted on three benchmark datasets, comparing our method with several classical methods. Evaluation results demonstrate that our proposed method captures semantic information more accurately and comprehensively. By effectively accomplishing automatic image annotation, our method can enhance human-machine-interaction to improve the level of intelligence in IIoT flexible manufacturing. © 2014 IEEE.