相关参考文献
注意:仅列出部分参考文献,下载原文获取全部文献信息。
Article
Computer Science, Artificial Intelligence
Rongli Gai et al.
Summary: Digital agriculture has a significant impact on the value of agricultural output. Robotic picking and cherry fruit detection technologies enable more intelligent and productive farming practices.
NEURAL COMPUTING & APPLICATIONS
(2023)
Article
Agronomy
Defang Xu et al.
Summary: To address the issues of fruit detection, such as ripeness and environmental conditions, a method called YOLO-Jujube was proposed. It incorporated various networks and loss functions to automatically detect jujube fruit and inspect its ripeness, outperforming other similar networks in terms of detection performance.
Article
Green & Sustainable Science & Technology
Poonam Dhiman et al.
Summary: In this study, an efficient model for citrus fruit disease prediction is proposed, which combines deep learning models CNN and LSTM with edge computing. By employing an enhanced feature-extraction mechanism and a feature-fusion subsystem, significant recognition on edge computing devices with retaining citrus fruit disease detection accuracy is achieved. The proposed CNN-LSTM model achieves an accuracy rate of 97.18% with Magnitude-Based Pruning and 98.25% with Magnitude-Based Pruning with Post Quantization, outperforming the existing CNN method.
Article
Agronomy
Jia Qin et al.
Summary: With advances in precision agriculture, autonomous agricultural machines can reduce human labor, optimize workflow, and increase productivity. In this study, deep learning was applied to LiDAR-based 3D obstacle detection in the farmland environment. By using deep learning algorithms, a data acquisition platform was established, and an effective 3D obstacle detector was trained. The proposed model significantly improves the detection performance for small objects and increases the average precision in the pedestrian class.
Article
Biochemistry & Molecular Biology
Hongzhi Song et al.
Summary: Cancer is a complex disease caused by genetic and nongenetic alterations. This paper proposes a novel semisupervised deep graph learning framework, GGraphSAGE, to predict cancer driver genes by integrating multiomics data. The experimental results show that GGraphSAGE outperforms other computational methods and provides insights into cancer driver genes specific to different tumor types. This research has significance in advancing precision medicine and predicting drivers for other complex diseases.
Article
Health Care Sciences & Services
Giorgia Marullo et al.
Summary: The study proposes a multi-task deep learning model for real-time blood accumulation detection and tools semantic segmentation in laparoscopic surgery videos. The system utilizes a shared backbone and separate branches to classify blood accumulation events and output segmentation maps. The multi-task approach achieves satisfactory results in surgical video tests, trained only with RGB images. The proposed network achieves an 81.89% Dice score for semantic segmentation and a 90.63% accuracy for event detection.
JOURNAL OF PERSONALIZED MEDICINE
(2023)
Article
Agronomy
Lanhui Fu et al.
Summary: This study proposes and compares two improved YOLOv4 neural network detection models, YOLO-Banana and YOLO-Banana-l4, in a banana orchard. The results show that both models can reduce network weight and shorten detection time compared to YOLOv4. The YOLO-Banana model has the best performance, with high detection accuracy for banana bunches and stalks in natural environment.
Review
Computer Science, Interdisciplinary Applications
Jan Egger et al.
Summary: Deep learning has made remarkable impact in various scientific disciplines, such as image processing and autonomous driving. It has also shown great potential in the medical domain. However, obtaining a complete overview of the field of 'medical deep learning' is becoming increasingly difficult due to the abundance of patient data and the rapid growth of deep learning research.
COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE
(2022)
Article
Agronomy
Pham Thi Quynh Anh et al.
Summary: The study introduced a modified multi-class model and a multi-label model for classifying garlic bulbs after root trimming. The multi-label model showed better overall accuracy in different classes. By adding a background image class, the classification accuracy of both models further improved.
POSTHARVEST BIOLOGY AND TECHNOLOGY
(2022)
Article
Agronomy
Fenghua Wang et al.
Summary: This paper proposes an improved YOLOv5s-CFL model for real-time detection of fruit targets in the Xiaomila green pepper picking robot. The model enhances detection accuracy and efficiency through modifications in the convolutional layer, the addition of a Coordinate Attention layer, and replacement of the Path Aggregation Network with the Bidirectional Feature Pyramid Network. Experimental results show that the improved model outperforms other models in detecting Xiaomila green peppers, and it has lightweight and real-time capabilities.
Article
Computer Science, Artificial Intelligence
Yi-Fan Zhang et al.
Summary: This paper focuses on the crucial step of bounding box regression in object detection and proposes two improvements, named EIOU loss and Focal-EIOU loss, to address the issues with previous loss functions. Experimental results demonstrate notable superiority in convergence speed and localization accuracy compared to other methods.
Article
Agronomy
Zhiyong Li et al.
Summary: The paper proposes an improved YOLOX target detection algorithm for accurately detecting the ripeness categories of sweet cherries in complex environments. By constructing the SweetCherry dataset and embedding the Convolutional Block Attention Module (CBAM) in the YOLOX model, the algorithm improves the model's attention to different channels and spaces, and replaces the original bounding box loss function with Efficient IoU (EIoU) loss for more accurate regression of prediction boxes. Experimental results show that the proposed method outperforms traditional algorithms and the YOLOX model in terms of various metrics, and can handle complex backgrounds and provide data and references for similar target detection problems.
Article
Agronomy
Fujie Zhang et al.
Summary: A real-time sorting robot system for classifying Panax notoginseng taproots based on the improved DeepLabv3+ model was developed. The system achieved high segmentation accuracy and detection speed, meeting the requirements of Panax notoginseng processing enterprises.
Article
Agriculture, Multidisciplinary
Hamzeh Mirhaji et al.
Summary: Fruit load estimation is crucial for Precision Agriculture. The study applied YOLO detection models to detect and count ripe oranges in an orchard, achieving high performance and accuracy through transfer learning and RGB image training. The YOLO-V4 model showed promising results in detecting and estimating orange yield, with high precision and performance.
COMPUTERS AND ELECTRONICS IN AGRICULTURE
(2021)
Article
Agricultural Engineering
P. Flores et al.
Summary: This study collected color RGB and color-infrared images of volunteer corn (VC) and soybean seedlings in a greenhouse, and found that support vector machine (SVM) and GoogLeNet performed best on the fused images. GoogLeNet was selected for its high accuracy, reasonable computation time, and ease of application.
INDUSTRIAL CROPS AND PRODUCTS
(2021)
Article
Chemistry, Analytical
Yinlong Zhu et al.
Summary: This study establishes a classification model for Panax notoginseng taproots based on image feature fusion, validating the importance of color, texture, and fusion features for classification through various models. Multiple dimensionality reduction methods are used to establish the main root classification model, providing a basis for developing online classification methods for different grades of Panax notoginseng in actual production.
Review
Agriculture, Multidisciplinary
Muhammad Hammad Saleem et al.
Summary: In recent years, there has been a growing interest in the application of automation in agriculture using artificial intelligence techniques and robotic systems. Significant improvements in agricultural tasks have been observed with the advancements in machine learning and deep learning concepts. Studies show that deep learning outperforms traditional machine learning algorithms in areas such as plant disease detection, crop recognition, and land cover classification.
PRECISION AGRICULTURE
(2021)
Article
Food Science & Technology
Limiao Deng et al.
Summary: This study developed an automatic carrot grading system based on computer vision and deep learning, utilizing a lightweight model and multiple techniques to detect and grade surface defects of carrots, achieving high accuracy and performance.
LWT-FOOD SCIENCE AND TECHNOLOGY
(2021)
Article
Computer Science, Artificial Intelligence
Jun Sun et al.
MACHINE VISION AND APPLICATIONS
(2020)
Article
Computer Science, Artificial Intelligence
Shaoqing Ren et al.
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
(2017)
Article
Plant Sciences
Chunlan Tang et al.
JOURNAL OF ETHNOPHARMACOLOGY
(2015)