Differential Feature Fusion and Gated Coordinate Attention Mechanism for Few‐Shot Object Detection of Electrical Equipment Images

Abstract

ABSTRACT The over‐dependence on the amount of labelled images in the process of deep learning‐based object detection of electrical equipment images leads to the consumption of manpower and material resources. To solve this problem, this paper proposes a few‐shot detection method of electrical equipment based on a two‐stage fine‐tuning approach and the improved YOLOv7 framework. First, an improved gated coordinate information attention (GCA) is introduced into the backbone part to reduce the negative impact of irrelevant background information on feature extraction. Second, in the neck part, an adaptive spatial feature fusion with difference information (ASFFDI) network is designed to improve the multi‐scale feature extraction ability. Finally, the optimised α‐SIoU loss function is used to improve the performance of the model. In addition, the two‐stage fine‐tuning training approach (TFA) is applied to the improved YOLOv7 model. The experimental results show that the average accuracy of the proposed method reaches 33.0%, 45.7% and 65.2%, respectively in 3 way‐10 shot, 3 way‐20 shot and 3 way‐30 shot, which can realise few‐shot detection of electrical equipment images.

Related Papers

• → A shot-by-shot breakdown of Remembrance(2019)1 cited
• → Study on viewer’s preference of sensibility vocabulary depending on composition of portrait shot in image -Mainly on the basis of size and placement of shot-(2021)1 cited
• → <title>High-speed photographic study on shot put</title>(1995)
• Study on Shot and be Shot Skills of Our Men Soccer Player from the World Cup(2002)
• → Take a Shot: Part 1(2017)