Robust detection of farmed fish by fusing YOLOv5 with DCM and ATM

Li, Haiqing; Hong, Yu; Gao, Haotian; Zhang, Peng; Wei, Sixue; Xu, Jingwen; Cheng, Siqi; Wu, Junfeng

doi:10.1016/j.aquaeng.2022.102301

Cited by 9 publications

(1 citation statement)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To solve the problem of a low recognition rate due to high aggregation of underwater organisms, Li et al [22] proposed CME-YOLOv5L, which introduces a CA attention mechanism to YOLOv5L to improve the loss function; this is better for dense fish detection. Li et al [23] proposed DCM-ATM-YOLOv5X, which uses a deformable convolution module (DCM) to extract fish features and an adaptive threshold module (ATM) to detect fish occlusion in Takifugu rubripes, but it still had a manually set threshold of 0.5. Zhao Meng et al [24] proposed a farmed fish detection SK-YOLOv5x model that fuses the SKNet (selective kernel networks) visual attention mechanism with YOLOv5x to form a feature extraction network focusing on pixel-level information, which enhances the recognition of fuzzy fish bodies.…”

Section: Related Workmentioning

confidence: 99%

Triple Attention Mechanism with YOLOv5s for Fish Detection

Long,

Wang,

et al. 2024

Fishes

View full text Add to dashboard Cite

Traditional fish farming methods suffer from backward production, low efficiency, low yield, and environmental pollution. As a result of thorough research using deep learning technology, the industrial aquaculture model has experienced gradual maturation. A variety of complex factors makes it difficult to extract effective features, which results in less-than-good model performance. This paper proposes a fish detection method that combines a triple attention mechanism with a You Only Look Once (TAM-YOLO)model. In order to enhance the speed of model training, the process of data encapsulation incorporates positive sample matching. An exponential moving average (EMA) is incorporated into the training process to make the model more robust, and coordinate attention (CA) and a convolutional block attention module are integrated into the YOLOv5s backbone to enhance the feature extraction of channels and spatial locations. The extracted feature maps are input to the PANet path aggregation network, and the underlying information is stacked with the feature maps. The method improves the detection accuracy of underwater blurred and distorted fish images. Experimental results show that the proposed TAM-YOLO model outperforms YOLOv3, YOLOv4, YOLOv5s, YOLOv5m, and SSD, with a mAP value of 95.88%, thus providing a new strategy for fish detection.

show abstract

Section: Related Workmentioning

confidence: 99%