检索结果-内蒙古大学图书馆

Semantic segmentation of track image based on deep neural network

The Journal of China Universities of Posts and Telecommunications 2020年第5期27卷 23-33页

作者： Wang Zhaoying Zhou Junhua Liao Zhonghua Zhai Xiang Zhang Lianping School of Electronic Engineering Beijing University of Posts and TelecommunicationsBeijing 100876China Beijing Complex Product Advanced Manufacturing Kngineering Research Center Beijing Simulation CenterBeijing 100854China State Key Laboratory of Intelligent Manufacturing System Technology Beijing Institute of Electronic System EngineeringBeijing 100854China Alibaba Cloud Computing Beijing 100020China

In this paper,deep learning technology was utilited to solve the railway track recognition in intrusion detection *** railway track recognition can be viewed as semantic segmentation task which extends image processing to pixel level *** encoder-decoder architecture DeepLabv3+model was applied in this work due to its good performance in semantic segmentation *** images of the railway track collected from the video surveillance of the train cab were used as experiment dataset in this work,the following improvements were made to the *** first aspect deals with over-fitting problem due to the limited amount of training *** augmentation and transfer learning are applied consequently to rich the diversity of data and enhance model robustness during the training ***,different gradient descent methods are compared to obtain the optimal optimizer for training model *** third problem relates to data sample imbalance,cross entropy(CE)loss is replaced by focal loss(FL)to address the issue of serious imbalance between positive and negative *** of the improved DeepLabv3+model with above solutions is demonstrated by experiment results with different system parameters.

关键词： railway track recognition convolutional neural networks semantic segmentation DeepLabv3+

来源：评论

学校读者我要写书评

暂无评论

Multi-scale Contrastive Learning for Gastroenteroscopy Classification

Multi-scale Contrastive Learning for Gastroenteroscopy Class...

引用

Annual IEEE Symposium on Computer-Based Medical systems

作者： Dan Li Xuechen Li Zhibin Peng Wenting Chen Linlin Shen Guangyao Wu Computer Vision Institute School of Computer Science & Software Engineering Shenzhen University National Engineering Laboratory for Big Data System Computing Technology ShenZhen University Shenzhen China City University of Hong Kong Hong Kong SAR China Shenzhen Institute of Artificial Intelligence & Robotics for Society Guangdong Key Laboratory of Intelligent Information Processing Shenzhen University General Hospital

In gastroenteroscopy image analysis, numerous CADs demonstrate that deep learning aids doctors' diagnosis. The shapes and sizes of the lesions are varied. And in the clinic, the dataset appears to be data imbalanced. However, existing methods directly classify by texture and ignore lesions with various shapes and sizes. To address the issue above, we propose a deep neural network, which consists of multi-scale feature extraction, contrastive feature learning and a multi-scale feature fusion module. We train the contrastive feature learning module and multi-scale feature fusion module simultaneously to alleviate the issue of data distribution differences. Thus, the proposed network can better identify various categories. Extensive experiments on the Hyper Kvasir dataset show that the proposed Hybrid-M2CL outperforms the benchmark proposed by the dataset with 5.0% Macro Precision, 3.3% Macro Recall, 3.4% Macro F1-score, 3.3% Micro Precision, 3.6% MCC. In addition, it outperforms the SOTA by 1.1% Macro F1-score, 2.6% MCC, and 2.0% B-ACC.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Provably Convergent Federated Trilevel Learning

arXiv

引用

arXiv 2023年

作者： Jiao, Yang Yang, Kai Wu, Tiancheng Jian, Chengtao Huang, Jianwei Department of Computer Science and Technology Tongji University China Key Laboratory of Embedded System and Service Computing Ministry of Education Tongji University China Shanghai Research Institute for Intelligent Autonomous Systems China School of Science and Engineering The Chinese University of Hong Kong Shenzhen China Shenzhen Institute of Artificial Intelligence and Robotics for Society China

Trilevel learning, also called trilevel optimization (TLO), has been recognized as a powerful modelling tool for hierarchical decision process and widely applied in many machine learning applications, such as robust neural architecture search, hyperparameter optimization, and domain adaptation. Tackling TLO problems has presented a great challenge due to their nested decision-making structure. In addition, existing works on TLO face the following key challenges: 1) they all focus on the non-distributed setting, which may lead to privacy breach;2) they do not offer any non-asymptotic convergence analysis which characterizes how fast an algorithm converges. To address the aforementioned challenges, this paper proposes an asynchronous federated trilevel optimization method to solve TLO problems. The proposed method utilizes µ-cuts to construct a hyper-polyhedral approximation for the TLO problem and solve it in an asynchronous manner. We demonstrate that the proposed µ-cuts are applicable to not only convex functions but also a wide range of non-convex functions that meet the µ-weakly convex assumption. Furthermore, we theoretically analyze the non-asymptotic convergence rate for the proposed method by showing its iteration complexity to obtain ϵ-stationary point is upper bounded by (Formula Presented). Extensive experiments on real-world datasets have been conducted to elucidate the superiority of the proposed method, e.g., it has a faster convergence rate with a maximum acceleration of approximately 80%. Copyright © 2023, The Authors. All rights reserved.

关键词： Decision making

来源：评论

学校读者我要写书评

暂无评论

Self-supervised GANs with label augmentation 21

Self-supervised GANs with label augmentation

引用

Proceedings of the 35th International Conference on Neural Information Processing systems

作者： Liang Hou Huawei Shen Qi Cao Xueqi Cheng Data Intelligence System Research Center Institute of Computing Technology Chinese Academy of Sciences and University of Chinese Academy of Sciences Data Intelligence System Research Center Institute of Computing Technology Chinese Academy of Sciences CAS Key Laboratory of Network Data Science and Technology Institute of Computing Technology Chinese Academy of Sciences and University of Chinese Academy of Sciences

ISBN: (纸本)9781713845393

Recently, transformation-based self-supervised learning has been applied to generative adversarial networks (GANs) to mitigate catastrophic forgetting in the discriminator by introducing a stationary learning environment. However, the separate self-supervised tasks in existing self-supervised GANs cause a goal inconsistent with generative modeling due to the fact that their self-supervised classifiers are agnostic to the generator distribution. To address this problem, we propose a novel self-supervised GAN that unifies the GAN task with the self-supervised task by augmenting the GAN labels (real or fake) via self-supervision of data transformation. Specifically, the original discriminator and self-supervised classifier are unified into a label-augmented discriminator that predicts the augmented labels to be aware of both the generator distribution and the data distribution under every transformation, and then provide the discrepancy between them to optimize the generator. Theoretically, we prove that the optimal generator could converge to replicate the real data distribution. Empirically, we show that the proposed method significantly outperforms previous self-supervised and data augmentation GANs on both generative modeling and representation learning across benchmark datasets.

关键词：

来源：评论

学校读者我要写书评

暂无评论

LCW: A Lightweight Recommendation Framework for Non-profit Crowdfunding Projects

LCW: A Lightweight Recommendation Framework for Non-profit C...

引用

作者： Jinxin Liu Yingyuan Xiao Wenguang Zheng Engineering Research Center of Learning-Based Intelligent System Ministry of EducationTianjin University of Technology Tianjin Key Laboratory of Intelligence Computing and Novel Software Technology Tianjin University of Technology

In recent years, crowdfunding platform has gradually emerged as a new economic model, and more and more people choose crowdfunding to obtain the initial funds needed by the project. However,with the increasing number of projects, it is difficult for sponsors to find suitable projects they want to invest in. Therefore, it is very important to build an efficient recommendation system on the crowdfunding platform. Different from previous crowdfunding recommendation systems designed to improve returns, this paper designs a lightweight recommendation framework LCW for nonprofit crowdfunding platforms. A large number of experiments on real kiva data sets have proved the effectiveness of this method and it obtains the SOTA results.

关键词： crowdfunding neural networks recommendation non-profit

来源：评论

学校读者我要写书评

暂无评论

Forensicability Assessment of Questioned Images in Recapturing Detection

arXiv

引用

arXiv 2022年

作者： Chen, Changsheng Zhao, Lin Cai, Rizhao Yu, Zitong Huang, Jiwu Kot, Alex C. The Guangdong Key Laboratory of Intelligent Information Processing and Shenzhen Key Laboratory of Media Security National Engineering Laboratory for Big Data System Computing Technology College of Electronics and Information Engineering Shenzhen University Shenzhen China Shenzhen Institute of Artificial Intelligence and Robotics for Society China The Department of Electrical and Electrics Engineering Nanyang Technological University Singapore639798 Singapore

Recapture detection of face and document images is an important forensic task. With deep learning, the performances of face anti-spoofing (FAS) and recaptured document detection have been improved significantly. However, the performances are not yet satisfactory on samples with weak forensic cues. The amount of forensic cues can be quantified to allow a reliable forensic result. In this work, we propose a forensicability assessment network to quantify the forensicability of the questioned samples. The low-forensicability samples are rejected before the actual recapturing detection process to improve the efficiency of recapturing detection systems. We first extract forensicability features related to both image quality assessment and forensic tasks. By exploiting domain knowledge of the forensic application in image quality and forensic features, we define three task-specific forensicability classes and the initialized locations in the feature space. Based on the extracted features and the defined centers, we train the proposed forensic assessment network (FANet) with cross-entropy loss and update the centers with a momentum-based update method. We integrate the trained FANet with practical recapturing detection schemes in face anti-spoofing (FAS) and recaptured document detection tasks. Experimental results show that, for a generic CNN-based FAS scheme, FANet reduces the EERs from 33.75% to 19.23% under ROSE to IDIAP protocol by rejecting samples with the lowest 30% forensicability scores. The performance of FAS schemes is poor in the rejected samples, with EER as high as 56.48%. Similar performances in rejecting low-forensicability samples have been observed for the state-of-the-art approaches in FAS and recaptured document detection tasks. To the best of our knowledge, this is the first work that assesses the forensicability of recaptured document images and improves the system efficiency. Copyright © 2022, The Authors. All rights reserved.

关键词： Image quality

来源：评论

学校读者我要写书评

暂无评论

A Physics-guided Multimodal Transformer Path to Weather and Climate Sciences

arXiv

引用

arXiv 2025年

作者： Han, Jing Chen, Hanting Han, Kai Huang, Xiaomeng Hu, Yongyun Xu, Wenjun Tao, Dacheng Zhang, Ping School of Artificial Intelligence Beijing University of Posts and Telecommunications Beijing China Huawei Noah’s Ark Lab Beijing China Department of Earth System Science Tsinghua University Beijing China Institute of Ocean Research Peking University Beijing China College of Computing & Data Science Nanyang Technological University Singapore State Key Laboratory of Networking and Switching Technology Beijing University of Posts and Telecommunications Beijing China

With the rapid development of machine learning in recent years, many problems in meteorology can now be addressed using AI models. In particular, data-driven algorithms have significantly improved accuracy compared to traditional methods. Meteorological data is often transformed into 2D images or 3D videos, which are then fed into AI models for learning. Additionally, these models often incorporate physical signals, such as temperature, pressure, and wind speed, to further enhance accuracy and interpretability. In this paper, we review several representative AI + Weather/Climate algorithms and propose a new paradigm where observational data from different perspectives, each with distinct physical meanings, are treated as multimodal data and integrated via transformers. Furthermore, key weather and climate knowledge can be incorporated through regularization techniques to further strengthen the model’s capabilities. This new paradigm is versatile and can address a variety of tasks, offering strong generalizability. We also discuss future directions for improving model accuracy and interpretability. Copyright © 2025, The Authors. All rights reserved.

关键词： Engineering education

来源：评论

学校读者我要写书评

暂无评论

A Fault Classification Prediction Model Based on Loose-Type Wavelet Neural Network for Server Health Management

A Fault Classification Prediction Model Based on Loose-Type ...

引用

Cloud computing, Big Data Applications and Software Engineering (CBASE), International Conference on

作者： Yi Wei Yingxiong Nong Zhe Li Jian Pan Zhenyu Yang Xiaoqi Lu Ningjiang Chen China Tobacco Guangxi Industrial Co. Ltd Nanning China School of Computer Electronics and Information Guangxi University Nanning China School of Computer Electronics and Information Guangxi University Guangxi Intelligent Digital Services Research Center of Engineering Technology Key Laboratory of Parallel Distributed and Intelligent Computing (Guangxi University) Nanning China

To prevent the serious impact on production efficiency caused by service interruption, data loss, and other issues resulting from server faults, this paper proposes a server fault prediction model named Wavelet Packet Transform Probabilistic Neural Network(WPT-PNN). WPT-PNN loosely mixes wavelet packet transform and probabilistic neural network to achieve quick fault localization and signal denoising effects. The proposed model is validated using server operational data gathered. The experimental results suggest that the WPT-PNN model can effectively manage the challenges of complicated, non-stationary, and noisy signals in the feature extraction stage and extract signal features reliably. In the fault classification prediction stage, our method improves fault prediction accuracy to 81%, and limits the error range within [-2, 4], better matching the requirements for precise and low false-positive fault prediction in servers.

关键词：

来源：评论

学校读者我要写书评

暂无评论

ACDR-CRAFF Net: A Multi-Scale Network Based on Adaptive Channel and Coordinate Relational Attention Network for Remote Sensing Scene Classification

引用

IET Image Processing 2025年第1期19卷

作者： Dai, Wei Xu, Haixia Shi, Furong Yuan, Liming Wang, Xinyu Wen, Xianbin School of Computer Science and Engineering Tianjin University of Technology Tianjin China Key Laboratory of Computer Vision and System Ministry of Education Tianjin University of Technology Tianjin China Tianjin Key Laboratory of Intelligence Computing and Novel Software Technology Tianjin University of Technology Tianjin China

Accurate classification of remote sensing scene images is crucial for diverse applications, from environmental monitoring to urban planning. While convolutional neural networks (CNNs) have dramatically improved classification accuracy, challenges remain due to the complex distribution of small objects, varied spatial configurations, and intra-class multimodality in remote sensing images. In this work, we make three key contributions to address these challenges. (1) We propose the adaptive channel and coordinate relational attention network (ACDR-CRAFF), a novel multi-scale feature fusion framework designed to enhance feature representation across scales. (2) We introduce two innovative modules: the adaptive channel dimensionality reduction (ACDR) module, which dynamically adjusts channel representations to retain essential low-dimensional features, and the coordinate relational attention multi-scale feature fusion (CRAFF) module, which effectively captures and transfers spatial information between feature levels. (3) By integrating ACDR and CRAFF, our model achieves a progressive fusion of local to global features, ensuring robust feature expressiveness at multiple scales. Experimental results on four widely used benchmark datasets demonstrate that ACDR-CRAFF consistently outperforms several state-of-the-art methods, achieving significant improvements in classification accuracy and setting a new benchmark for complex remote sensing scene classification tasks. These results underscore the effectiveness of our approach in addressing the limitations of existing methods and advancing the state of the art in remote sensing image analysis. © 2025 The Author(s). IET Image Processing published by John Wiley & Sons Ltd on behalf of The Institution of Engineering and technology.

关键词： Photointerpretation

来源：评论

学校读者我要写书评

暂无评论

A Mother-Son Model for Multi-Objective Non-Invasive Load Monitoring

SSRN

引用

SSRN 2023年

作者： Mei, Huawei Liu, Yanping Cao, Wangbin Yu, Yang North China Electric Power University Baoding071003 China Engineering Research Center of Intelligent Computing for Complex Energy Systems Ministry of Education Baoding071003 China Hebei Key Laboratory of Power Internet of Things Technology North China Electric Power University Baoding071003 China State Key Laboratory of Alternate Electrical Power System with Renewable Energy Sources North China Electric Power University Baoding071003 China

This paper proposes a mother-son model (MSM) to address the challenges of missing data, complex models, and limited transfer capabilities in non-intrusive load monitoring (NILM). The MSM model designs a novel transfer method based on feature inheritance to dynamically decompose multiple appliances (including unknown appliances). The mother model incorporates a six-parallel feature extraction structure to extract comprehensive features, which are then inherited by the son model. The features inherited from the mother model integrate information from multiple dimensions in the spatial domain and focus on local fine features in the temporal domain. The son model offers the capability to adjust the number of appliances and decompose unknown appliances, catering to the requirements of practical applications of NILM. Furthermore, a novel statistical metric, Ratio Squared Error (RSE), is designed to quantify the proportion of squared error in energy consumption. The experiments are carried out on five households and statistically evaluated using five statistical indicators. The results indicate that the son model exhibits superior decomposition performance and greater resilience compared to the state-of-the-art model. © 2023, The Authors. All rights reserved.

关键词： Energy utilization

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：