检索结果-内蒙古大学图书馆

Development of an Optimized YOLO-PP-Based Cherry Tomato Detection System for Autonomous Precision Harvesting

PROCESSES 2025年第2期13卷 353-353页

作者： Qin, Xiayang Cao, Jingxing Zhang, Yonghong Dong, Tiantian Cao, Haixiao Nanjing Univ Informat Sci & Technol Sch Automat Nanjing 210044 Peoples R China Wuxi Siasun Robot & Automat Co Ltd Wuxi 214101 Peoples R China

An accurate and efficient detection method for harvesting is crucial for the development of automated harvesting robots in short-cycle, high-yield facility tomato cultivation environments. This study focuses on cherry tomatoes, which grow in clusters, and addresses the complexity and reduced detection speed associated with the current multi-step processes that combine target detection with segmentation and traditional image processing for clustered fruits. We propose YOLO-Picking Point (YOLO-PP), an improved cherry tomato picking point detection network designed to efficiently and accurately identify stem keypoints on embedded devices. YOLO-PP employs a C2FET module with an EfficientViT branch, utilizing parallel dual-path feature extraction to enhance detection performance in dense scenes. Additionally, we designed and implemented a Spatial Pyramid Squeeze Pooling (SPSP) module to extract fine features and capture multi-scale spatial information. Furthermore, a new loss function based on Inner-CIoU was developed specifically for keypoint tasks to further improve detection *** model was tested on a real greenhouse cherry tomato dataset, achieving an accuracy of 95.81%, a recall rate of 98.86%, and mean Average Precision (mAP) scores of 99.18% and 98.87% for mAP50 and mAP50-95, respectively. Compared to the DEKR, YOLO-Pose, and YOLOv8-Pose models, the mAP value of the YOLO-PP model improved by 16.94%, 10.83%, and 0.81%, respectively. The proposed algorithm has been implemented on NVIDIA Jetson edge computing devices, equipped with a human-computer interaction interface. The results demonstrate that the proposed Improved Picking Point Detection Network exhibits excellent performance and achieves real-time accurate detection of cherry tomato harvesting tasks in facility agriculture.

关键词： keypoint detection YOLO v8 tomato detection facility agriculture attention mechanism deep learning

来源：评论

学校读者我要写书评

暂无评论

Improved Transunet's Segmentation Study of Bone Scintigraphy Lesions 2

Improved Transunet's Segmentation Study of Bone Scintigraphy...

引用

2nd International Conference on Machine Vision, image processing and Imaging Technology, MVIPIT 2024

作者： Cao, Rui Luo, Renze School of Electrical and Information Southwest Petroleum University Chengdu China School of Faculty of Earth Sciences and Technology Southwest Petroleum University Chengdu China

ISBN: (纸本)9798331543037

In order to solve the problems of irregular targets and fuzzy boundaries in bone scintigraphy segmentation, an improved TransUNet model was proposed. The feature extraction part of the encoder is replaced with an asymmetric convolution residual module to enhance feature capture in different directions and avoid gradient vanishing. At the same time, the cross-fusion module is used to replace the jump connection, which strengthens the deep connection between the encoder and the decoder, suppresses redundant information and improves fine-grained feature capture. In addition, the maximization decision-making method in the two-channel output dimension is used to improve the richness of classification information, capture the uncertain region and reduce the influence of class imbalance in the case of fuzzy boundaries, and obtain a segmentation graph..Experimental results show that the improved Transunet segmentation algorithm has improved the Intersection over Union (loU), DSC coefficient (Dice Similarity Cofficient), pixel accuracy (CPA) and recall rate (Recall), reaching 0.498, 0.667, 0.662 and 0.761, respectively, which is better than the current mainstream segmentation algorithms. It has certain practical application value. ©2024 IEEE.

关键词： deep learning

来源：评论

学校读者我要写书评

暂无评论

PROAUG: PROTOTYPE-BASED AUGMENTATION FOR LONG-TAILED image CLASSIFICATION 49

PROAUG: PROTOTYPE-BASED AUGMENTATION FOR LONG-TAILED IMAGE C...

引用

49th IEEE International Conference on Acoustics, Speech, and Signal processing (ICASSP)

作者： Hong, Yan Zhang, Jianfu Sun, Zhongyi Yan, Ke Ant Grp Hangzhou Peoples R China Shanghai Jiao Tong Univ Shanghai Peoples R China Youtu Lab Tencent Beijing Peoples R China

ISBN: (纸本)9798350344868;9798350344851

real-world data often exhibit long-tailed distributions with heavy class imbalance, which deteriorates the generalization performance of the classifier. To mitigate this problem, we propose a novel Prototype-based Augmentation framework (ProAug) to address the data scarcity issue by augmenting the feature space for tail classes. Our ProAug consists of a prototype construction branch and a dynamic augmentation branch. The prototype-based dictionary is optimized with category-aware margin loss to learn multi-center and discriminative prototypes for each category. In the dynamic augmentation branch, we aim to produce high-quality tail-class features by dynamically composing context-similar prototypes with an attention mechanism. Moreover, to further improve the reliability of prototypes and the quality of augmented features, a meta-update strategy is adopted to calibrate two branches of ProAug to boost performance. Extensive empirical results on CIFAR-LT-10/100, imageNet-LT, and iNaturalist 2018 demonstrate the effectiveness of our method.

关键词： Long-tail classification margin loss prototype meta-update strategy deep learning

来源：评论

学校读者我要写书评

暂无评论

EasyClick: A New Web-based image Annotation Tool for Person Re-identification Using deep learning 24

EasyClick: A New Web-based Image Annotation Tool for Person ...

引用

6th International Conference on image, Video and Signal processing, IVSP 2024

作者： Bashar, Khayrul Fujimoto, Yusei Kaneoka, Yuichiro R and D Team Intelligence Design Inc. Tokyo Japan

ISBN: (纸本)9798400716829

image annotation is a vital step for model building and object recognition. Although fully automatic annotation is expected, it still has limitations in the scenario like person re-identification (ReID) where multi-camera images are involved. This difficulty arises due to complex intra-class variations in illumination, pose, viewpoint, blur, low resolution, and occlusion. Researchers typically use manual annotation tools that annotate gallery images by searching them one-by-one for each query image, which is a tedious and time-consuming task. In the study, we propose a new web-based semi-automatic annotation tool, called "EasyClick", which capitalizes the capacity of a deep learning model, called omni-scale network (OSNet), with cosine similarity metric and a clustering algorithm. Our proposed approach is a versatile one that can provide two ranking suggestions with and without using a hierarchical clustering algorithm. Given a query image from one camera, this tool can retrieve a small subset of the most similar images or a few clusters of ranked images from another camera. Users can then select all relevant images to a query by easily clicking the displayed images. Several experiments with two datasets having 43,246 (802 persons) and 594 (124 persons) multi-camera images showed promising performance of the proposed tool in terms of speed and accuracy when compared to the popular CVAT annotation tool. © 2024 ACM.

关键词： image annotation

来源：评论

学校读者我要写书评

暂无评论

SecFePAS: Secure Facial-expression-based Pain Assessment with deep learning at the Edge 9

SecFePAS: Secure Facial-expression-based Pain Assessment wit...

引用

9th Symposium on Edge Computing

作者： Batool, Kanwal Anwar, Saleem Mann, Zoltan Adam Univ Amsterdam Amsterdam Netherlands Rotterdam Univ Appl Sci Rotterdam Netherlands Univ Halle Wittenberg Halle Germany

ISBN: (纸本)9798350378290;9798350378283

Patient monitoring in hospitals, nursing centers, and home care can be largely automated using cameras and machine-learning-based video analytics, thus considerably increasing the efficiency of patient care. In particular, Facial-expression-based Pain Assessment Systems (FePAS) can automatically detect pain and notify medical personnel. However, current FePAS solutions using cloud-based video analytics offer very limited security and privacy protection. This is problematic, as video feeds of patients constitute highly sensitive information. To address this problem, we introduce SecFePAS, the first FePAS solution with strong security and privacy guarantees. SecFePAS uses advanced cryptographic protocols to perform neural network inference in a privacy-preserving way. To counteract the significant overhead of the used cryptographic protocols, SecFePAS uses multiple optimizations. First, instead of a cloud-based setup, we use edge computing with a 5G connection to benefit from lower network latency. Second, we use a combination of transfer learning and quantization to devise neural networks with high accuracy and optimized inference time. Third, SecFePAS quickly filters out unessential frames of the video to focus the in-depth analysis on key frames. We tested SecFePAS with the SqueezeNet and ResNet50 neural networks on a real pain estimation benchmark. SecFePAS outperforms state-of-the-art FePAS systems in accuracy and optimizes secure processing time.

关键词： deep learning edge computing facial-expression-based pain assessment homomorphic encryption multi-party computation neural networks privacy-preserving machine learning privacy-preserving protocols secure computation secure inference secure video classification

来源：评论

学校读者我要写书评

暂无评论

Integrated Aquaculture Monitoring System Using Combined Wireless Sensor Networks and deep Reinforcement learning

引用

Sensors and Materials 2024年第3期36卷 1019-1033页

作者： Sung, Wen-Tsai Isa, Indra Griha Tofik Hsiao, Sung-Jung Department of Electrical Engineering National Chin-Yi University of Technology Zhongshan Rd Section 2 No. 57 Taichung City411030 Taiwan Department of Information Technology Takming University of Science and Technology Taipei City11451 Taiwan

Freshwater fish is one of the commodities experiencing an increasing growth rate from 1990 to 2018. Many efforts have been made to meet market needs, through both fisheries technology and applied technology, one of which is an integrated monitoring system. In this study, an aquaculture monitoring system was developed that integrates wireless sensor networks (WSNs) based on temperature, pH, and turbidity with deep reinforcement learning. The purpose of this study is to produce a convenient, precise, and low-cost aquaculture monitoring system. The stages of the study are (1) the integration of all the WSN components, (2) the validation of the WSNs, (3) the implementation of the analysis model in the system, (4) the implementation of the recommended model into the DRL system, and (5) practical experimentation using the aquaculture monitoring system. The WSN validation results indicate that the average percentage error is 3.23%, whereas at the system modeling stage, the optimal accuracy is 98.80%. In the experiment to monitor real aquaculture environmental conditions, an accuracy of 97% is obtained. © 2024 M Y U Scientific Publishing Division. All rights reserved.

关键词： Wireless sensor networks

来源：评论

学校读者我要写书评

暂无评论

Augmenting Clinical Decisions with deep learning Lung Cancer image Abnormality Segmentation 14

Augmenting Clinical Decisions with Deep Learning Lung Cancer...

引用

14th International Conference on Cloud Computing, Data Science and Engineering, Confluence 2024

作者： Venkatraman, K. Reddy, Sirigiri Naga Pavan Sathvik Amrita Vishwa Vidyapeetham School of Computing Computer Science and Engineering Chennai India

ISBN: (纸本)9798350344837

Lung cancer is a major global cause of death, highlighting the critical need for quick and accurate detection methods. The exploration of computational alternatives arose from the standard way of manually processing CT images, which is time-consuming and error-prone. In this work, we combine the advantages of Support Vector Machine (SVM) and VGG16 classifiers to provide a novel method for enhancing lung cancer diagnosis. By analyzing the 'IQ-OTH/NCCD' dataset, our hybrid model-which combines the VGG 16 and SVM algorithms-performs admirably in differentiating between aggressive, benign, and normal lung cancer cases. This combination of traditional machine learning with deep learning tackles accuracy and efficiency issues, which is a promising development over current diagnostic methods. We conduct a comprehensive comparative analysis with prominent architectures to select the optimal model based on accuracy, efficiency, and resource requirements. In addition to introducing the VGG 16+SVM model, our research provides valuable insights into deep learning architectures, with the ultimate goal of advancing precise and efficient diagnosis of lung cancer, which is crucial in combating this global health challenge in the future. © 2024 IEEE.

关键词： Computerized tomography

来源：评论

学校读者我要写书评

暂无评论

Research on Automobile Intelligent Manufacturing Defect Detection Algorithm Based on deep learning 4

Research on Automobile Intelligent Manufacturing Defect Dete...

引用

4th IEEE International Conference on Data Science and Computer Application, ICDSCA 2024

作者： Lv, Yuanyuan Geng, Xiao Zhang, Xiaorong Dai, Jiarong Shandong Vocational and Technical University of Engineering Jinan Shandong China Jinan Yiheng Technology Co. Ltd Jinan Shandong China

ISBN: (纸本)9798350368239

This research puts forward a deep-learning-centered automotive manufacturing defect detection algorithm. It utilizes the SSD (Single Shot MultiBox Detector) algorithm to realize the efficient detection of surface flaws on automotive components. Initially, the system employs CNN to extract image features and combines multi-scale features for the purpose of strengthening the recognition capability of diverse defects. Subsequently, the model incorporates automatic tagging and data augmentation techniques to enhance the generalization ability of the model with respect to different defect types. This research has designed a range of experimental scenarios and has verified the effectiveness of the algorithm through accurate data analysis. The experimental outcomes indicate that the proposed algorithm has attained a high level in terms of accuracy, recall rate and other metrics. In comparison with traditional detection methods, it exhibits greater robustness and real-time performance. Precisely, in the actual test set, the algorithm has achieved a detection accuracy of 95.6% and a recall rate of 92.8%, which has effectively enhanced the detection efficiency and decreased the false detection rate. The research findings demonstrate that the defect detection algorithm based on deep learning has extensive application prospects in the automotive intelligent manufacturing field, and is anticipated to significantly boost the automation level of the manufacturing process and the reliability of product quality. ©2024 IEEE.

关键词： Smart manufacturing

来源：评论

学校读者我要写书评

暂无评论

Mitigating Cyberbullying in Social Media: A deep Contextual learning Approach for Severity Level Classification in Textual Data 5

Mitigating Cyberbullying in Social Media: A Deep Contextual ...

引用

5th International Conference on Electronics and Sustainable Communication Systems, ICESC 2024

作者： Agrawal, Prashant Kumar, Awanit Tripathi, Arun Kr. Sangam University Computer Science & Engineering Rajasthan India Sangam University Bhilwara Computer Science & Engineering Rajasthan India Kiet Group of Institution Department of Computer Applications Delhi-NCR Ghaziabad India

ISBN: (纸本)9798350379945

This research introduces a novel approach that integrates deep Contextual learning (DCL), specifically the DCL-256-32 model with an embedding model to accurately classify offense levels within the textual data. The DCL-256-32 model employs a SoftMax function to assign probabilities to distinct severity classes, ranging from critical to negligible. The proposed model incorporates two endpoints: an embedding model for generating semantic representations of input text and a set of pre-trained DCL-256-32 models for predicting offense levels. By averaging these predictions and associating them with humanreadable labels, this study proposes a robust and scalable framework for real-time text analysis. The proposed model demonstrates high performance compared to existing methods, contributing to the advancement of Natural Language processing (NLP) and classification. This research study offers a practical solution for enhancing digital safety and combating online harassment. © 2024 IEEE.

关键词： Contrastive learning

来源：评论

学校读者我要写书评

暂无评论

deep learning-Powered Face Detection and Recognition for Challenging Environments 2

Deep Learning-Powered Face Detection and Recognition for Cha...

引用

2nd International Conference on Intelligent Data Communication Technologies and Internet of Things, IDCIoT 2024

作者： Reddy, Thiyyagura Jagadeesh Ganesh, Marrey Sai Kumar Reddy, Mandadi Hemanth Bhandhavya, Chintharapu Jansi, R. SRM Institute of Science and Technology SRM Nagar Faculty of Engineering and Technology Department of Electronics and Communicati on Engineering Kattankulathur603203 India

ISBN: (纸本)9798350327533

The increasing prevalence of surveillance systems in both public and private domains underscores the growing need for robust human face detection and recognition capabilities. This research introduces an innovative real-time framework that leverages deep learning techniques, particularly Convolutional Neural Networks (CNNs), to accurately detect human faces in complex images. Through a combination of image preprocessing, training using diverse dataset and classification using trained model, this system excels at identifying faces even in challenging scenarios marked by low lighting and diverse angles. This capability enables the precise identification of individuals of interest. Rigorous testing on publicly available dataset showcases the system's outstanding performance in terms of face detection and recognition accuracy. To validate the outstanding performance of CNN is face detection, comparison was done using other machine learning algorithms like k-NN, random forest and logistic regression. This research contributes to the advancement of facial recognition technology, offering a reliable solution for video surveillance applications. It holds the potential to enhance security, surveillance, and forensic practices, benefiting stakeholders ranging from law enforcement agencies to public safety organizations and private security firms. © 2024 IEEE.

关键词： Accuracy CNN deep learning Facial Recognition Machine learning

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：