An Improved Few-Shot Object Detection via Feature Reweighting Method for Insulator Identification

被引:3
|
作者
Wu, Junpeng [1 ,2 ]
Zhou, Yibo [2 ]
机构
[1] Northeast Elect Power Univ, Key Lab Modern Power Syst Simulat & Control & Rene, Minist Educ, Jilin 132012, Peoples R China
[2] Northeast Elect Power Univ, Sch Elect Engn, Jilin 132012, Peoples R China
来源
APPLIED SCIENCES-BASEL | 2023年 / 13卷 / 10期
关键词
insulator; few-shot object detection; feature extraction network; attention mechanism; non-maximum suppression algorithm;
D O I
10.3390/app13106301
中图分类号
O6 [化学];
学科分类号
0703 ;
摘要
To address the issue of low accuracy in insulator object detection within power systems due to a scarcity of image sample data, this paper proposes a method for identifying insulator objects based on improved few-shot object detection through feature reweighting. The approach utilizes a meta-feature transfer model in conjunction with the improved YOLOv5 network to realize insulator recognition under conditions of few-shot. Firstly, the feature extraction module of the model incorporates an improved self-calibrated feature extraction network to extract feature information from multi-scale insulators. Secondly, the reweighting module integrates the SKNet attention mechanism to facilitate precise segmentation of the mask. Finally, the multi-stage non-maximum suppression algorithm is designed in the prediction layer, and the penalty function about confidence is set. The results of multiple prediction boxes are retained to reduce the occurrence of false detection and missing detection. For the poor detection results due to a low diversity of sample space, the transfer learning strategy is applied in the training to transfer the entire trained model to the detection of insulator targets. The experimental results show that the insulator detection mAP reaches 29.6%, 36.0%, and 48.3% at 5-shot, 10-shot, and 30-shot settings, respectively. These findings serve as evidence of improved accuracy levels of the insulator image detection under the condition of few shots. Furthermore, the proposed method enables the recognition of insulators under challenging conditions such as defects, occlusion, and other special circumstances.
引用
收藏
页数:18
相关论文
共 50 条
  • [41] A Closer Look at Few-Shot Object Detection
    Liu, Yuhao
    Dong, Le
    He, Tengyang
    PATTERN RECOGNITION AND COMPUTER VISION, PRCV 2023, PT VIII, 2024, 14432 : 430 - 447
  • [42] Few-Shot Object Detection with Weight Imprinting
    Yan, Dingtian
    Huang, Jitao
    Sun, Hai
    Ding, Fuqiang
    COGNITIVE COMPUTATION, 2023, 15 (05) : 1725 - 1735
  • [43] Spatial reasoning for few-shot object detection
    Kim, Geonuk
    Jung, Hong-Gyu
    Lee, Seong-Whan
    PATTERN RECOGNITION, 2021, 120
  • [44] Few-Shot Object Detection via Classify-Free RPN
    Yu, Songlin
    Yang, Zhiyu
    Zhang, Shengchuan
    Cao, Liujuan
    PATTERN RECOGNITION AND COMPUTER VISION, PRCV 2023, PT X, 2024, 14434 : 101 - 112
  • [45] Few-shot object detection via data augmentation and distribution calibration
    Zhu, Songhao
    Zhang, Kai
    MACHINE VISION AND APPLICATIONS, 2024, 35 (01)
  • [46] Few-Shot Object Detection via Classification Refinement and Distractor Retreatment
    Li, Yiting
    Zhu, Haiyue
    Cheng, Yu
    Wang, Wenxin
    Teo, Chek Sing
    Xiang, Cheng
    Vadakkepat, Prahlad
    Lee, Tong Heng
    2021 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION, CVPR 2021, 2021, : 15390 - 15398
  • [47] Few-shot object detection via data augmentation and distribution calibration
    Songhao Zhu
    Kai Zhang
    Machine Vision and Applications, 2024, 35
  • [48] Industrial few-shot fractal object detection
    Huang, Haoran
    Luo, Xiaochuan
    Yang, Chen
    NEURAL COMPUTING & APPLICATIONS, 2023, 35 (28): : 21055 - 21069
  • [49] Few-Shot Object Detection with Model Calibration
    Fan, Qi
    Tang, Chi-Keung
    Tai, Yu-Wing
    COMPUTER VISION, ECCV 2022, PT XIX, 2022, 13679 : 720 - 739
  • [50] Hallucination Improves Few-Shot Object Detection
    Zhang, Weilin
    Wang, Yu-Xiong
    2021 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION, CVPR 2021, 2021, : 13003 - 13012