检索结果-内蒙古大学图书馆

ADeepWeeD: An adaptive deep learning framework for weed species classification

学校读者我要写书评

暂无评论

Artificial Intelligence in Agriculture

Artificial Intelligence in Agriculture 2025年第4期15卷 590-609页

作者： Rahman, Md Geaur Rahman, Md Anisur Parvez, Mohammad Zavid Patwary, Md Anwarul Kaium Ahamed, Tofael Fleming-Muñoz, David A. Aloteibi, Saad Moni, Mohammad Ali School of Computing Mathematics and Engineering Charles Sturt University Port MacquarieNSW2444 Australia La Trobe Business School La Trobe University MelbourneVIC3086 Australia School of Physics Maths and Computing The University of Western Australia PerthWA6009 Australia Faculty of Life and Environmental Sciences University of Tsukuba Tsukuba Science City Ibaraki305-8572 Japan Computer Science Department Community College King Saud University Riyadh11437 Saudi Arabia AI & Digital Health Technology Artificial Intelligence and Cyber Futures Institute Charles Sturt University BathurstNSW2795 Australia AI & Digital Health Technology Rural Health Research Institute Charles Sturt University OrangeNSW2800 Australia School of IT Washington University of Science and Technology VA United States La Trobe Institute for Sustainable Agriculture and Food La Trobe University MelbourneVIC3086 Australia

Efficient weed management in agricultural fields is essential for attaining optimal crop yields and safeguarding global food security. Every year, farmers worldwide invest significant time, capital, and resources to combat yield losses, approximately USD 75.6 billion, due to weed infestations. Deep Learning (DL) methodologies have been recently implemented to revolutionise agricultural practices, particularly in weed detection and classification. Existing DL-based weed classification techniques, including VGG16 and ResNet50, initially construct a model by implementing the algorithm on a training dataset comprising weed species, subsequently employing the model to identify weed species acquired during training. Given the dynamic nature of crop fields, we argue that existing methods may exhibit suboptimal performance due to two key issues: (i) the unavailability of all training weed species initially, as these species emerge over time, resulting in a progressively expanding training dataset, and (ii) the constrained memory and computational capacity of the system utilised for model development, which hinders the retention of all weed species that manifest over an extended duration. To address the issues, this paper introduces a novel DL-based framework called ADeepWeeD for weed classification that facilitates adaptive (i.e. incremental) learning so that it can handle new weed species by keeping track of historical information. ADeepWeeD is evaluated using two criteria, namely F1-Score and classification accuracy, by comparing its performances against four non-incremental and two incremental state-of-the-art methods on three publicly available large datasets. Our experimental results demonstrate that ADeepWeeD outperforms existing techniques used in this study. We believe that our developed model could be used to develop an automation system for weed identification. The code of the proposed method is available on GitHub: https://***/grahman20/ADeepWeed. © 2025 T

关键词： Weed control

MiM: Mask in Mask Self-Supervised Pre-Training for 3D Medical Image Analysis

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Zhuang, Jiaxin Wu, Linshan Wang, Qiong Fei, Peng Vardhanabhuti, Varut Luo, Lin Chen, Hao Departments of Computer Science and Engineering Hong Kong University of Science and Technology Hong Kong Shenzhen Institutes of Advanced Technology Chinese Academy of Sciences China School of Optical and Electronic Information Wuhan National Laboratory for Optoelectronics Huazhong University of Science and Technology Wuhan430074 China Department of Diagnostic Radiology The University of Hong Kong Hong Kong College of Engineering Peking University Beijing China Department of Computer Science and Engineering Department of Chemical and Biological Engineering State Key Laboratory of Molecular Neuroscience Hong Kong University of Science and Technology Hong Kong HKUST Shenzhen-Hong Kong Collaborative Innovation Research Institute Futian Shenzhen China

The Vision Transformer (ViT) has demonstrated remarkable performance in Self-Supervised Learning (SSL) for 3D medical image analysis. Masked AutoEncoder (MAE) for feature pre-training can further unleash the potential of ViT on various medical vision tasks. However, due to large spatial sizes with much higher dimensions of 3D medical images, the lack of hierarchical design for MAE may hinder the performance of downstream tasks. In this paper, we propose a novel Mask in Mask (MiM) pre-training framework for 3D medical images, which aims to advance MAE by learning discriminative representation from hierarchical visual tokens across varying scales. We introduce multiple levels of granularity for masked inputs from the volume, which are then reconstructed simultaneously ranging at both fine and coarse levels. Additionally, a cross-level alignment mechanism is applied to adjacent level volumes to enforce anatomical similarity hierarchically. Furthermore, we adopt a hybrid backbone to enhance the hierarchical representation learning efficiently during the pre-training. MiM was pre-trained on a large scale of available 3D volumetric images, i.e., Computed Tomography (CT) images containing various body parts. Extensive experiments on twelve public datasets demonstrate the superiority of MiM over other SSL methods in organ/tumor segmentation and disease classification. We further scale up the MiM to large pre-training datasets with more than 10k volumes, showing that large-scale pre-training can further enhance the performance of downstream tasks. Code and checkpoint will be available once upon acceptance. Copyright © 2024, The Authors. All rights reserved.

关键词： computerized tomography

Technical Support for Detection and Prediction of Rainfall 2

学校读者我要写书评

暂无评论

Technical Support for Detection and Prediction of Rainfall

2nd International Conference on Smart Electronics and Communication, ICOSEC 2021

作者： Hemavathi, D. Kumar, V. Ramesh Regin, R. Rajest, S. Suman Phasinam, Khongdet Singh, Saurabh SRM Institute of Science and Technology School of Computing Department of Information Technology Tamil Nadu Physical Education and Sports University Department of Sports Management Chennai India Adhiyamaan College of Engineering Department of Information Technology Tamil Nadu India Technology Advanced Studies Vels Institute of Science Tamil Nadu Chennai India Pibulsongkram Rajabhat University School of Agricultural and Food Engineering Faculty of Food and Agricultural Technology Phitsanulok Thailand Prestige Institute of Management and Research India

ISBN: (纸本)9781665433686

This project evaluates the effectiveness of some DM methods for predicting rainfall using India's historical weather data gathered from the IDM database. Including 10 attributes from the collected meteorological data, only 5 features are generally believed to be significant for precipitation forecasting. The following phases of research are used in the collected data: data cleaning, data selection, Data Transformation and classification. Many pre-processing responsibilities are concerned with determining missing values and eliminating clutter. Hence, two responsibilities are carrying out for finding missing value purpose, which is cleaning and normalization. In this work, six data mining classification methods, including MLR, K-NN, SVM, ANN, Random Forest (RF) and decision tree (DT), were investigated. The classification process divides objects into different classes. A classifier is often trained using a training set, where single or multiple experts assign labels to a set of objects. The proposed RF is a novel classification technique used in rainfall data to classify the region differently from normal rainfall or heavy rainfall. The completed accuracy of the planned RF classifier is 96.1%, which is advanced than other classifiers used in the Rainfall database. © 2021 IEEE.

关键词： Rain

Privacy-preserving Decision Making Based on Q-Learning in Cloud computing

学校读者我要写书评

暂无评论

Privacy-preserving Decision Making Based on Q-Learning in Cl...

IEEE International Conference on Trust, Security and Privacy in computing and Communications (TrustCom)

作者： Zhipeng Zhou Chenyu Dong Donger Mo Peijia Zheng School of Computer Science and Engineering Sun Yat-sen University Guangzhou China School of Computer Science & Engineering GuangDong Province Key Laboratory of Information Security Technology Sun Yat-sen University Guangzhou China Zhengzhou Xinda Institute of Advanced Technology Zhengzhou China

People encounter a variety of continuous decision-making (DM) problems in the real world. Reinforcement learning (RL) is a promising technique to solve these problems. This paper proposes a privacy-preserving Q-learning decision-making scheme (PQDM). Based on distributed homomorphic encryption (HE), we design several secure protocols to implement the under-lying nonlinear operations such as comparing, maximizing, and maximizing parameter solving. Based on the designed security protocols, we propose a secure decision-making protocol in cloud computing, which enables the cloud server to perform element selection and Q-learning functions on ciphertext data. During the entire process, the cloud server does not need to know the actual state, thus guaranteeing the security of the original state information. We analyze the security and complexity of the whole scheme theoretically. Our experimental results show our proposed scheme’s effectiveness and good spatio-temporal performance.

关键词： Cloud computing Privacy Protocols Q-learning Computational modeling Decision making Data models

Query-guided Prototype Evolution Network for Few-Shot Segmentation

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Cong, Runmin Xiong, Hang Chen, Jinpeng Zhang, Wei Huang, Qingming Zhao, Yao Institute of Information Science Beijing Jiaotong University Beijing100044 China School of Control Science and Engineering Shandong University Jinan250061 China Key Laboratory of Machine Intelligence and System Control Ministry of Education Jinan250061 China Beijing Key Laboratory of Advanced Information Science and Network Technology Beijing100044 China Department of Computer Science City University of Hong Kong Hong Kong School of Computer Science and Technology University of Chinese Academy of Sciences Beijing101408 China

Previous Few-Shot Segmentation (FSS) approaches exclusively utilize support features for prototype generation, neglecting the specific requirements of the query. To address this, we present the Query-guided Prototype Evolution Network (QPENet), a new method that integrates query features into the generation process of foreground and background prototypes, thereby yielding customized prototypes attuned to specific queries. The evolution of the foreground prototype is accomplished through a support-query-support iterative process involving two new modules: Pseudo-prototype Generation (PPG) and Dual Prototype Evolution (DPE). The PPG module employs support features to create an initial prototype for the preliminary segmentation of the query image, resulting in a pseudo-prototype reflecting the unique needs of the current query. Subsequently, the DPE module performs reverse segmentation on support images using this pseudo-prototype, leading to the generation of evolved prototypes, which can be considered as custom solutions. As for the background prototype, the evolution begins with a global background prototype that represents the generalized features of all training images. We also design a Global Background Cleansing (GBC) module to eliminate potential adverse components mirroring the characteristics of the current foreground class. Experimental results on the PASCAL-5i and COCO-20i datasets attest to the substantial enhancements achieved by QPENet over prevailing state-of-the-art techniques, underscoring the validity of our ideas. © 2024, CC BY-NC-SA.

关键词： Semantic Segmentation

Graph Confidence Intercalibration Network for Intracranial Aneurysm Lesion Instance Segmentation in Dsa

学校读者我要写书评

暂无评论

SSRN

SSRN 2024年

作者： Ye, Haili Mo, Yancheng Tang, Chen Liao, Mingqian Zhang, Xiaoqing Dai, Limeng Li, Baihua Liu, Jiang Research Institute of Trustworthy Autonomous Systems Department of Computer Science and Engineering Southern University of Science and Technology Shenzhen518055 China Department of Computer Science Loughborough University LoughboroughLE11 3TU United Kingdom School of Ophthalmology and Optometry Eye Hospital Wenzhou Medical University Wenzhou325027 China Center for High Performance Computing and Shenzhen Key Laboratory of Intelligent Bioinformatics Shenzhen Institute of Advanced Technology Chinese Academy of Sciences Shenzhen518055 China Shenzhen People’s Hospital Neurosurgery Department Shenzhen518020 China China

Intracranial aneurysm (IA) lesion segmentation is significant for its treatment and prognosis. Although exiting deep network-based instance methods have good IA lesion segmentation results based on digital subtraction angiography (DSA) images, they still face great challenges with instance confidence bias and imprecise boundary segmentation, which may negatively affect IA diagnosis. To tackle these problems, this paper proposes a novel graph confidence intercalibration network (GCINet) to automatically segment IA lesions from DSA images. To be specific, we design a graph confidence intercalibration (GCI) module to mitigate instance confidence bias by dynamically adjusting their confidence distributions. At the same time, we propose an edge space perception (ESP) module to correct ambiguous segmentation boundaries. Extensive experiments on a clinical IA-DSA and a publicly available LiTS dataset demonstrate that our GCINet outperforms state-of-the-art methods. Additionally, visual analysis and ablation studies are provided to verify the effectiveness of each module in GCINet. © 2024, The Authors. All rights reserved.

关键词： Image segmentation

Regularly Truncated M-estimators for Learning with Noisy Labels

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Xia, Xiaobo Lu, Pengqian Gong, Chen Han, Bo Yu, Jun Yu, Jun Liu, Tongliang The Sydney AI Center School of Computer Science Faculty of Engineering The University of Sydney DarlingtonNSW2008 Australia The Australian AI Institute Faculty of Engineering and IT The University of Technology Sydney BroadwayNSW2007 Australia The PCA Lab The Key Laboratory of Intelligent Perception and Systems for High-Dimensional Information of Ministry of Education School of Computer Science and Engineering Nanjing University of Science and Technology Nanjing210094 China The Department of Computing Hong Kong Polytechnic University Hong Kong The Department of Computer Science Hong Kong Baptist University Hong Kong The School of Computer Science and Technology Hangzhou Dianzi University Hangzhou310018 China The Department of Automation University of Science and Technology of China Hefei230026 China

The sample selection approach is very popular in learning with noisy labels. As deep networks "learn pattern first", prior methods built on sample selection share a similar training procedure: the small-loss examples can be regarded as clean examples and used for helping generalization, while the large-loss examples are treated as mislabeled ones and excluded from network parameter updates. However, such a procedure is arguably debatable from two folds: (a) it does not consider the bad influence of noisy labels in selected small-loss examples;(b) it does not make good use of the discarded large-loss examples, which may be clean or have meaningful information for generalization. In this paper, we propose regularly truncated M-estimators (RTME) to address the above two issues simultaneously. Specifically, RTME can alternately switch modes between truncated M-estimators and original M-estimators. The former can adaptively select small-losses examples without knowing the noise rate and reduce the side-effects of noisy labels in them. The latter makes the possibly clean examples but with large losses involved to help generalization. Theoretically, we demonstrate that our strategies are label-noise-tolerant. Empirically, comprehensive experimental results show that our method can outperform multiple baselines and is robust to broad noise types and levels. The implementation is available at https://***/xiaoboxia/RTM_LNL. Copyright © 2023, The Authors. All rights reserved.

关键词： Machine learning

Analysis of the Requirement and Artificial Intelligence-based Resource Management System in Cloud

学校读者我要写书评

暂无评论

Analysis of the Requirement and Artificial Intelligence-base...

International Conference on advanced computing and Communication Systems (ICACCS)

作者： S. Nagasundaram B. N. Bobinath Ashwini Shedthi K. Rajalakshmi Trupti Dandekar Humnekar Millerjothi School of Computing Sciences Vels Institute of Science Technology and Advanced Studies Chennai TamilNadu India Department of Computer Science and Engineering NPR College of Engineering Dindigul TamilNadu India Department of Commerce and Management New Horizon College Marathahalli and Research Scholar at Amrita University Bangalore India Department of Marketing CMS Business School Jain (Deemed to be) University Bangalore India Department of Mechanical Sri Eshwar College of Engineering Coimbatore India

Artificial Intelligence (AI) approaches, such as the big data, mobile computing, and Internet of Things (IoT) have become more common in the computer science area in recent years. It is critical to properly manage resources in order to maintain service quality, service level agreements, and the system's overall availability. AI may be used to address the needs of cloud resource management, and this study examines and evaluates several approaches. Fog computing systems, intelligent cloud computing systems and edge-cloud systems are all types of AI-based cloud resource management systems that have been reviewed in this contribution. In this research, an intelligent resource management method that controls mobile resources by monitoring device states and projecting their future stability is proposed. We look at the possibility of using our suggested resource management system to a variety of cloud-based platforms.

关键词： Cloud computing Stability analysis Mobile handsets Virtual machining Resource management Internet of Things Reliability

Solving Aspect category sentiment analysis as a text generation task

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Liu, Jian Teng, Zhiyang Cui, Leyang Liu, Hanmeng Zhang, Yue School of Computer Science Fudan University School of Engineering Westlake University Institute of Advanced Technology Westlake Institute for Advanced Study

Aspect category sentiment analysis has attracted increasing research attention. The dominant methods make use of pre-trained language models by learning effective aspect category-specific representations, and adding specific output layers to its pre-trained representation. We consider a more direct way of making use of pre-trained language models, by casting the ACSA tasks into natural language generation tasks, using natural language sentences to represent the output. Our method allows more direct use of pre-trained knowledge in seq2seq language models by directly following the task setting during pre-training. Experiments on several benchmarks show that our method gives the best reported results, having large advantages in few-shot and zero-shot settings. © 2021, CC BY-NC-ND.

关键词： Sentiment analysis