Tighte: A Model for Campus Security Target Tracking in Edge Intelligent Computing Architecture DOI
Feng Zhou, Xin Zhao, Jing Liu

et al.

Published: June 28, 2024

Language: Английский

Typical Marine Ecological Disasters in China Attributed to Marine Organisms and Their Significant Insights DOI Creative Commons
Lulu Yao, Peimin He, Zhangyi Xia

et al.

Biology, Journal Year: 2024, Volume and Issue: 13(9), P. 678 - 678

Published: Aug. 30, 2024

Owing to global climate change or the ever-more frequent human activities in offshore areas, it is highly probable that an imbalance ecosystem has been induced. However, importance of maintaining and protecting marine ecosystems' balance cannot be overstated. In recent years, various disasters have occurred frequently, such as harmful algal blooms (green tides red tides), storm surge disasters, wave sea ice tsunami disasters. Additionally, overpopulation certain organisms (particularly faunas) led threatening both ecosystems safety. The ecological disaster monitoring system China primarily focuses on controlling outbreak green (mainly caused by outbreaks some

Language: Английский

Citations

6
A lightweight network based on SCMYOLO for accurate and efficient underwater fish detection DOI Creative Commons
Zhongde Zhang,

Qiaoling Han,

Weiping Liu

et al.

ICES Journal of Marine Science, Journal Year: 2025, Volume and Issue: 82(4)

Published: April 1, 2025

Abstract The accurate and rapid identification of fish in natural underwater settings is crucial for the intelligent monitoring fisheries. However, computational requirements intricate models necessitate advanced hardware capabilities, limiting deployment within systems. To address this, we introduce a lightweight object detection model named SCMYOLO. This employs several innovations to reduce complexity enhance performance. Firstly, replace yolov5 backbone network with MobileNetV3, significantly reducing parameters load as measured by floating-point operations. Secondly, integrate high-level screening-feature fusion pyramid structure Neck layer combine C3 module context guided (CGNetwork) module. approach improves extraction low-level features addresses challenges associated multi-scale detection, thereby enabling feature fusion. Lastly, incorporate SIoU loss function refine accuracy bounding box predictions. Our experimental findings demonstrate that SCMYOLO achieves an 88.4% reduction 86.2% decrease operations, size file constituting just 14.6% original. Furthermore, sustains high-performance metrics precision, recall, [email protected] all exceeding 90%, indicating robust capabilities. Given these attributes, particularly suited on various edge computing devices, offering efficient solution complex environments. Source code available at https://github.com/systemavailable/SCMYOLO.

Language: Английский

Citations

0
YOLOv7-GCM: a detection algorithm for creek waste based on improved YOLOv7 model DOI
Jianhua Qin,

Honglan Zhou,

Huaian Yi

et al.

Pattern Analysis and Applications, Journal Year: 2024, Volume and Issue: 27(4)

Published: Sept. 17, 2024

Language: Английский

Citations

1
Tighte: A Model for Campus Security Target Tracking in Edge Intelligent Computing Architecture DOI
Feng Zhou, Xin Zhao, Jing Liu

et al.

Published: June 28, 2024

Language: Английский

Citations

0