APM-YOLOv7 for Small-Target Water-Floating Garbage Detection Based on Multi-Scale Feature Adaptive Weighted Fusion

Abstract

As affected by limited information and the complex background, the accuracy of small-target water-floating garbage detection is low. To increase the detection accuracy, in this research, a small-target detection method based on APM-YOLOv7 (the improved YOLOv7 with ACanny PConv-ELAN and MGA attention) is proposed. Firstly, the adaptive algorithm ACanny (adaptive Canny) for river channel outline extraction is proposed to extract the river channel information from the complex background, mitigating interference of the complex background and more accurately extracting the features of small-target water-floating garbage. Secondly, the lightweight partial convolution (PConv) is introduced, and the partial convolution-efficient layer aggregation network module (PConv-ELAN) is designed in the YOLOv7 network to improve the feature extraction capability of the model from morphologically variable water-floating garbage. Finally, after analyzing the limitations of the YOLOv7 network in small-target detection, a multi-scale gated attention for adaptive weight allocation (MGA) is put forward, which highlights features of small-target garbage and decreases missed detection probability. The experimental results showed that compared with the benchmark YOLOv7, the detection accuracy in the form of the mean Average Precision (mAP) of APM-YOLOv7 was improved by 7.02%, that of mmAP (mAP0.5:0.95) was improved by 3.91%, and Recall was improved by 11.82%, all of which meet the requirements of high-precision and real-time water-floating garbage detection and provide reliable reference for the intelligent management of water-floating garbage.

Enhanced floating debris detection algorithm based on CDW-YOLOv8

“…Additionally, they further enhanced detection accuracy through strategies such as data augmentation, test-time augmentation, and weighted box fusion. Jiang et al [35] aimed to improve the precision in detecting small floating debris on water surfaces by incorporating lightweight partial convolution (PConv) technology and a multi-scale gate attention adaptive weight distribution (MGA) method into the YOLOv7 algorithm, resulting in the development of the APM-YOLOv7 algorithm tailored for this purpose.…”

A lightweight fine-grained recognition algorithm based on object detection

“…Sensors 23. https://doi.org/10.3390/s23198072) introduced Triple Attention to PSPNet's pyramid module, facilitating cross-dimensional interaction between spatial information to focus on specific positions of fish body features within channels and enhancing clarity around fish edge positions.Jiang et al (Jiang T, Zhou J, Xie BB et al (2024) Improved YOLOv8 Model for Lightweight Pigeon Egg Detection. Animals 14. https://doi.org/10.3390/ani14081226) also made significant contributions: they incorporated an EMA module into the C2f module and replaced the upsampling module in the neck network with Dysample when designing YOLOv8-PG for rapid detection of fragile egg-shaped objects; experimental results demonstrated substantial improvement from these modifications.Furthermore, Jiang et al (Jiang ZJ, Wu BJ, Ma L et al (2024) APM-YOLOv7 for Small-Target Water-Floating Garbage Detection Based on Multi-Scale Feature Adaptive Weighted Fusion. Sensors 24. https://doi.org/10.3390/s24010050) designed APM-YOLOv7 based on the YOLOv7 network, introduced lightweight convolution PConv and multi-scale gated attention for adaptive weight allocation (MGA), significantly enhancing its ability to extract features from various shapes of floating garbage on water surfaces-essentially meeting highprecision real-time requirements for detecting water surface floating garbage.The flexibility and effectiveness exhibited by attention mechanisms compared to deep CNN backbone architectures have positioned them as crucial components within computer vision research (Vaswani et al, Vaswani A, Shazeer N, Parmar N et al (2017) Attention Is All You Need.…”