CIOD: an intelligent class-incremental object detection system with nearest mean of exemplars

被引：0
作者：
Ren S. [1 ,2 ]
He Y. [2 ]
Wang X. [3 ]
Guo K. [2 ]
Barra S. [4 ]
Li J. [1 ]
机构：
[1] School of Computer and Electrical Engineering, Hunan University of Arts and Science, Dongting Avenue, Hunan, Changde
[2] School of Computer Science and Engineering, Central South University, Lushan South Road, Hunan, Changsha
[3] Department of Computer Science, St. Francis Xavier University, University Avenue, 1-902-863-3300, Antigonish, NS
[4] Department, University of Naples “Federico II”, Via Claudio 21, Campania, Naples
来源：
Journal of Ambient Intelligence and Humanized Computing | 2023年 / 14卷 / 08期
基金：
中国国家自然科学基金; 中国国家社会科学基金;
关键词：
Class-incremental; Deep learning; Intelligent system; Object detection;
D O I：
10.1007/s12652-022-04341-7
中图分类号：
学科分类号：
摘要：
Object detection has been widely used in intelligent video surveillance, robot navigation, industrial detection, and other fields. Object detection can effectively reduce the consumption of human capital and has important practical significance. However, the existing object detection methods lack adaptability to the application environment and need to obtain all images classes at one time to train the model in a static setting, and do not support the incremental learning. Therefore, we propose a novel Class-Incremental Object Detection (CIOD) framework based on deep learning. CIOD divides object detection into two stages: object candidate box generation and selection. In the first stage, we improved the traditional OpenCV cascaded classifier to adapt to class-incremental learning while maintaining accuracy. In the second stage, we use example sets and prototype vectors to construct a class increment-based classifier to identify the generated object candidate box. We verify the effectiveness of the proposed method in terms of object detection effect, efficiency, and memory capacity. Experimental results show that CIOD can detect object in a class-incremental manner and can control the memory capacity not to increase with the number of newly increased classes. © 2022, The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature.
引用
收藏
页码：10657 / 10671
页数：14
相关论文
共 38 条
  • [21] Mohammed M., Azzeddine D., Youssef B., Taoufiq G., Semmdpref: Algorithm to filter and sort rules using a semantically based ontology technique, Proceedings of the 7Th International Conference on Management of Computational and Collective Intelligence in Digital Ecosystems, pp. 29-34, (2015)
  • [22] Mohammed M.A., Abdulkareem K.H., Mostafa S.A., Et al., Voice pathology detection and classification using convolutional neural network model, Appl Sci, 10, 11, (2020)
  • [23] Rebuffi S.A., Kolesnikov A., Sperl G., Lampert C.H., Icarl: Incremental classifier and representation learning, Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 2001-2010, (2017)
  • [24] Redmon J., Farhadi A., Yolo9000: Better, faster, stronger, Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 7263-7271, (2017)
  • [25] Redmon J., Farhadi A., Yolov3: An incremental improvement, (2018)
  • [26] Redmon J., Divvala S., Girshick R., Farhadi A., You only look once: Unified, real-time object detection, Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 779-788, (2016)
  • [27] Ren S., He K., Girshick R., Sun J., Faster r-cnn: Towards real-time object detection with region proposal networks, Adv Neural Inf Process Syst, 28, (2015)
  • [28] Salim F.D., Loke S.W., Rakotonirainy A., Krishnaswamy S., U &i aware: A framework using data mining and collision detection to increase awareness for intersection users, 21St International Conference on Advanced Information Networking and Applications Workshops (AINAW’07), pp. 530-535, (2007)
  • [29] Shen F., Yang Y., Liu L., Liu W., Tao D., Shen H.T., Asymmetric binary coding for image search, IEEE Trans Multimedia, 19, 9, pp. 2022-2032, (2017)
  • [30] Viola P., Jones M., Rapid object detection using a boosted cascade of simple features. In: Proceedings of the 2001 IEEE computer society conference on computer vision and pattern recognition. CVPR 2001, Ieee, Pp I–I, (2001)
1234