U-YOLOv7: A network for underwater organism detection

Yu, Guoyan; Cai, Ruilin; Su, Jie; Hou, Mingxin; Deng, Ruoling

doi:10.1016/j.ecoinf.2023.102108

Cited by 23 publications

(4 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Specific improvements include the creation of a new backbone network, the adoption of a new Anchor-Free detection head, and the utilization of a new loss function. The backbone network refers to the ELAN design concept of YOLOv7 [11], using the gradient flow-rich C2f structure and adjusting the number of channels according to different models, signifi-cantly improving the model performance. The head part adopts the current mainstream decoupled head structure, separates the classification and detection heads, and the Anchor-Based is replaced by the Anchor-Free.…”

Section: Yolov8 Object Detection Algorithmmentioning

confidence: 99%

“…While these methods perform better in detection accuracy and localization precision, they come at the cost of slower detection speed. In contrast, one-stage algorithms such as YOLO [8][9][10][11] and SSD [12] transform the object detection and border localization problem directly into a single regression problem, thus eliminating the step of candidate box generation, greatly saving time and providing a significant advantage in detection speed to meet the tasks with strong real-time requirements. Although the single-stage algorithm may be slightly inferior to the two-stage algorithm in terms of detection accuracy, related research is rapidly advancing, and many scholars are working to improve the accuracy of the single-stage algorithm [13,14].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Sonar Fish School Detection and Counting Method Based on Improved YOLOv8 and BoT-SORT

Xing,

Sun,

Liu

et al. 2024

JMSE

View full text Add to dashboard Cite

Fish object detection and counting in pelagic fisheries face many challenges in complex environments. Sonar imaging technology offers a solution because it generates high-resolution images underwater. In this paper, we propose a sonar-based fish object detection and counting method using an improved YOLOv8 combined with BoT-SORT to address issues such as missed detection, false detection, and low accuracy caused by complex factors such as equipment motion, light changes, and background noise in pelagic environments. The algorithm utilizes the techniques of lightweight upsampling operator CARAFE, generalized feature pyramid network GFPN, and partial convolution. It integrates with the BoT-SORT tracking algorithm to propose a new region detection method that detects and tracks the schools of fish, providing stable real-time fish counts in the designated area. The experimental results indicate that while focusing on maintaining a lightweight design, the improved algorithm achieved a 3.8% increase in recall and a 2.4% increase in compared to the original algorithm. This significantly impacts scientific and rational fishery planning, marine resource protection, and improved productivity. At the same time, it provides important data support for marine ecological monitoring, environmental protection, and fishery management, contributing to sustainable fishery development and marine ecology preservation.

show abstract

Section: Yolov8 Object Detection Algorithmmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Sonar Fish School Detection and Counting Method Based on Improved YOLOv8 and BoT-SORT

Xing,

Sun,

Liu

et al. 2024

JMSE

View full text Add to dashboard Cite

show abstract

“…A new algorithm with stronger feature extraction capability is proposed. Yu et al [16] introduced 3D attention mechanism and structured the network with CrossConv and efficient squeezing excitation module based on YOLOv7 to improve the detection performance of the algorithm. Jia et al [17] added the InceptionNeXT module to the backbone network based on the YOLOv8n algorithm and added the attention module SEAM to the Neck section, the proposed algorithm has better detection accuracy underwater.…”

Section: ⅱRelated Workmentioning

confidence: 99%

Research on Underwater Small Target Detection Technology Based on Single-Stage USSTD-YOLOv8n

Yi,

Yang,

Yan

2024

IEEE Access

View full text Add to dashboard Cite

Aiming at the problem of low visibility of underwater environment, which leads to the leakage of small target detection and low accuracy, this paper proposes an improved algorithm USSTD-YOLOv8n (Underwater small-size target detection YOLOv8n) based on YOLOv8n. First, CARAFE is adopted as anew up-sampling method to achieve more correct feature reconstruction under low underwater visibility. Second, Context Guided Block (CG block) is introduced to replace part of the convolutional structure, which makes USSTD-YOLOv8n have stronger feature extraction capability. Finally, Inner-CIoU is adopted as the loss function to improve the generalization ability of USSTD-YOLOv8n, to obtain more correct detection results. To verify the robustness and accuracy of the model, a new experimental strategy is used to perform one set of ablation experiments and three sets of comparison experiments on the URPC2018 and URPC2020 datasets, the

show abstract

“…While traditional techniques face challenges, the application of convolutional neural networks (CNNs), notably YOLOv7, in honey pollen identification remains largely unexplored. Despite its proven efficacy in diverse fields, including agriculture, , food, aquaculture, transportation, and various other object detection tasks, , YOLOv7’s potential in honey-related applications lacks substantiated evidence. Our study aims to bridge this gap by investigating the viability of YOLOv7 for honey pollen detection, highlighting the need for a robust data set and providing guidance for future research in this emerging field.…”

Section: Introductionmentioning

confidence: 99%

Detection and Identification of Honey Pollens by YOLOv7: A Novel Framework toward Honey Authenticity

Jubayer,

Ruhad,

Kayshar

et al. 2024

ACS Agric. Sci. Technol.

View full text Add to dashboard Cite

Honey, a valuable and globally consumed food product, has significant market potential linked to its origin. However, authenticating honey is challenging due to sophisticated adulteration techniques. This current research introduces an innovative approach employing YOLOv7, a cutting-edge object detection model, to detect and classify honey pollens, thereby bolstering the authentication of honey. Our methodology involved creating a data set comprising three well-known honey varieties (Sundarban, Litchi, and Mustard), supplemented by three sets of unidentified honey pollen images sourced from Kaggle. Subsequently, we assembled a data set consisting of 3000 images representing the pollen types extracted from the known honey samples. To tackle the challenge of limited sample sizes, we employed data augmentation techniques. The efficacy of our approach was evaluated using established statistical measures including detection accuracy, precision, recall, mAP value, and F1 score, yielding impressive values of 98.3, 99.3, 100, 99.2%, and 0.985, respectively. The YOLOv7 model's reliability was validated using Kaggle's unknown honey pollen data sets, which showed that it correctly detected and identified these new pollens based on previous training. Through rigorous experimentation and validation, our study underscores the potential of the YOLOv7 framework in revolutionizing quality control practices within the honey industry, ensuring consumers access to genuine and top-tier honey products through pollen image analysis.

show abstract

U-YOLOv7: A network for underwater organism detection

Cited by 23 publications

References 32 publications

Sonar Fish School Detection and Counting Method Based on Improved YOLOv8 and BoT-SORT

Sonar Fish School Detection and Counting Method Based on Improved YOLOv8 and BoT-SORT

Research on Underwater Small Target Detection Technology Based on Single-Stage USSTD-YOLOv8n

Detection and Identification of Honey Pollens by YOLOv7: A Novel Framework toward Honey Authenticity

Contact Info

Product

Resources

About