检索结果-南通市图书馆

FSCIL-EACA: Few-Shot Class-Incremental Learning Network Based on Embedding Augmentation and Classifier Adaptation for Image Classification

引用

Chinese Journal of Electronics 2024年第1期33卷 139-152页

作者： Ruru ZHANG Haihong E meina song School of Computer Science Beijing University of Posts and Telecommunications Education Department Information Network Engineering Research Center Beijing University of Posts and Telecommunications

The ability to learn incrementally is critical to the long-term operation of AI systems. Benefiting from the power of few-shot class-incremental learning(FSCIL), deep learning models can continuously recognize new classes with only a few samples. The difficulty is that limited instances of new classes will lead to overfitting and exacerbate the catastrophic forgetting of the old classes. Most previous works alleviate the above problems by imposing strong constraints on the model structure or parameters, but ignoring embedding network transferability and classifier adaptation(CA), failing to guarantee the efficient utilization of visual features and establishing relationships between old and new classes. In this paper, we propose a simple and novel approach from two perspectives: embedding bias and classifier bias. The method learns an embedding augmented(EA) network with cross-class transfer and class-specific discriminative abilities based on self-supervised learning and modulated attention to alleviate embedding bias. Based on the adaptive incremental classifier learning scheme to realize incremental learning capability,guiding the adaptive update of prototypes and feature embeddings to alleviate classifier bias. We conduct extensive experiments on two popular natural image datasets and two medical datasets. The experiments show that our method is significantly better than the baseline and achieves state-of-the-art results.

关键词： Few-shot class-incremental learning Embedding augmentation Classifier adaptation Image classification

来源：

同方期刊数据库评论

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

FGM-SPCL:Open-Set Recognition Network for Medical Images Based on Fine-Grained Data Mixture and Spatial Position Constraint Loss

引用

Chinese Journal of Electronics 2024年第4期33卷 1023-1033页

作者： Ruru ZHANG Haihong E Lifei YUAN Yanhui WANG Lifei WANG meina song School of Computer Science Beijing University of Posts and Telecommunications Engineering Research Center of Information Networks Ministry of Education Hebei Eye Hospital

The current intelligent auxiliary diagnosis models all follow the closed-set recognition setting. After the model is deployed online, the input data is often not completely controlled. Diagnosing an untrained disease as a known category would lead to serious medical malpractice. Therefore, realizing the open-set recognition is significant to the safe operation of the intelligent auxiliary diagnosis model. Currently, most open-set recognition models are studied for natural images, and it is very challenging to obtain clear and concise decision boundaries between known and unknown classes when applied to fine-grained medical images. We propose an open-set recognition network for medical images based on fine-grained data mixture and spatial position constraint loss(FGM-SPCL) in this work.Considering the fine graininess of medical images and the diversity of unknown samples, we propose a fine-grained data mixture(FGM) method to simulate unknown data by performing a mixing operation on known data to expand the coverage of unknown data difficulty levels. In order to obtain a concise and clear decision boundary, we propose a spatial position constraint loss(SPCL) to control the position distribution of prototypes and samples in the feature space and maximize the distance between known classes and unknown classes. We validate on a private ophthalmic OCT dataset, and extensive experiments and analyses demonstrate that FGM-SPCL outperforms state-of-the-art models.

关键词：

来源：

同方期刊数据库评论

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Dynamically modulated synthesis of hollow metal-organic frameworks for selective hydrogenation reactions

引用

Nano Research 2023年第8期16卷 11334-11341页

作者： Liyun Xiao Chuanqi Cheng Zhixi Li Chaoyang Zheng Jing Du meina song Yue Wan Shaopeng Li Guo Jun Meiting Zhao Department of Chemistry Institute of Molecular Aggregation ScienceTianjin Key Laboratory of Molecular Optoelectronic SciencesTianjin UniversityTianjin 300072China Institute of New Energy Materials School of Materials Science and EngineeringTianjin UniversityTianjin 300072China State Key Laboratory of Separation Membranes and Membrane Processes School of ChemistryTiangong UniversityTianjin 300387China

Hollow metal-organic frameworks(MOFs)have attracted increasing attention in the field of catalysis in recent years due to their unique cavity structure with fast mass-diffusion rates and easily accessible active sites.Here,we report the use of dynamic modulators,which are formed by the in-situ imine condensation reaction of 4-aminobenzoic acid and 4-formylbenzoic acid,to regulate the growth of MOFs to synthesize well-defined hollow thioether functionalized UiO-67(denoted as H-UiO-67-S)single crystals.After supporting Pd nanoparticles,the designed catalysts Pd@H-UiO-67-S show excellent conversion(>99.9%),selectivity(>99.9%),and stability(10 cycles)in the selective hydrogenation of nitrobenzenes with other reducible groups.Density functional theory calculations and the experimental results reveal that Pd nanoparticles not only selectively adsorb the nitro-groups on nitrobenzene,but also restrict the adsorption of the aniline product,due to the interaction of thioether with Pd in the confined pores of H-UiO-67-S,finally result in a significant increase in selectivity of nitro-hydrogenation.

关键词： hollow structure metal-organic frameworks(MOFs) dynamically modulated synthesis selective hydrogenation nitrobenzene

来源：

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Traffic Prediction in 3G Mobile Networks Based on Multifractal Exploration

引用

Tsinghua Science and Technology 2013年第4期18卷 398-405页

作者： Yanhua Yu meina song Yu Fu Junde song the School of Computer Beijing University of Posts and Telecommunications

Traffic prediction plays an integral role in telecommunication network planning and network optimization. In this paper, we investigate the traffic forecasting for data services in 3G mobile networks. Although the Box-Jenkins model has been proven to be appropriate for voice traffic （since the arrival of calls follows a Poisson distribution）, it has been demonstrated that the Internet traffic exhibits statistical self-similarity and has to be modeled using the Fractional AutoRegressive Integrated Moving Average （FARIMA） process. However, a few studies have concluded that the FARIMA process may fail in modeling the Internet traffic. To this end, we conducted experiments on the modeling of benchmark Internet traffic and found that the FARIMA process fails because of the significant multifractal characteristic inherent in the traffic series. Thereafter, we investigate the traffic series of data services in a 3G mobile network from a province in China. Rich multifractal spectra are found in this series. Based on this observation, an integrated method combining the AutoRegressive Moving Average （ARMA） and FARIMA processes is applied. The obtained experimental results verify the effectiveness of the integrated prediction method.

关键词： time series prediction self-similar Fractional AutoRegressive Integrated Moving Average （FARIMA）

来源：

维普期刊数据库

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

ERCS: An Efficient and Robust Card Recognition System for Camera-Based Image

引用

China Communications 2020年第12期17卷 247-264页

作者： Zhonghong Ou Baiqiao Xiong Fenrui Xiao meina song Beijing University of Posts and Telecommunications No.10 Xitucheng Road Haidian DistrictBeijing 100876China

Cards Recognition Systems,(CRSs)are representative computer vision-based applications.They have a broad range of usage scenarios.For example,they can be used to recognize images containing business cards,personal identification cards,and bank cards etc.Even though CRSs have been studied for many years,it is still difficult to recognize cards in camera-based images taken by ordinary devices,e.g.,mobile phones.Diversity of viewpoints and complex backgrounds in the images make the recognition task challenging.Existing systems employing traditional image processing schemes are not robust to varied environment,and are inefficient in dealing with natural images,e.g.,taken by mobile phones.To tackle the problem,we propose a novel framework for card recognition by employing a Convolutional Neutral Network(CNN)based approach.The system localizes the foreground of the image by utilizing a Fully Convolutional Network(FCN).With the help of the foreground map,the system localizes the corners of the card region and employs perspective transformation to alleviate the effects from distortion.Text lines in the card region are detected and recognized by utilizing CNN and Long Short Term Memory,(LSTM).To evaluate the proposed scheme,we collect a large dataset which contains 4,065 images in a variety of shooting scenarios.Experimental results demonstrate the efficacy of the proposed scheme.Specifically,it is able to achieve an accuracy of 90.62%in the end-toend test,outperforming the state-of-the-art.

关键词： card localization card recognition optical character recognition CNN

来源：

维普期刊数据库

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

Research Progress of Volatile Substances Analysis and Sensory Evaluation of Soymilk

引用

Journal of Northeast Agricultural University(English Edition) 2024年第1期31卷 74-82页

作者： Xu Qiongyuan Li Ka Huang meina Liang Yuxin Tao Bingjie Zhang Tingting song Nannan Ji Yazhou Zhang Huajiang Xia Ning College of Food Science Northeast Agricultural UniversityHarbin 150030China Key Laboratory of Soybean Biology Northeast Agricultural UniversityHarbin 150030China

Soybean is one of the important crops in China. Soymilk, a traditional neutral plant-based protein drink, is rich in high quality proteins. Although soybean milk is rich in nutrients, its marketing among consumers, especially those in Western countries who are used to peaceful flavor, has been limited due to the adverse flavor impact brought by its special composition. In recent years, with the increasing attention to the nutritional value of soymilk, the flavor of soymilk has become a popular research object for scholars at home and abroad. The flavor components of soymilk are mainly volatile small molecular compounds produced by enzymatic reactions catalyzed by lipoxygenase(LOX). After formation, they interact with protein macromolecules to form the overall flavor of soymilk. At present, there are many methods to control the off-odor of soymilk at home and abroad, including physical heating methods, chemical methods, biological enzymatic digestion methods, mask methods, and a variety of breeding methods. These methods effectively reduce the off-odor of soymilk, but all of them have shortcomings. Currently, the sensory characteristics of the beany odor in soymilk are evaluated mainly by traditional human sensory scoring along with the assistance of modern instrument analysis of volatile flavor substances using headspace solid phase microextraction(SPME) gas chromatography coupled with-mass spectrometry(GC-MS). This paper summarized the research results of volatile flavor substances in soymilk in recent years and the sensory evaluation methods of soymilk at home and abroad, and looked forward to the future development direction, hoping to provide some theoretical bases and reference detection methods for solving the problem of soymilk flavor in the future.

关键词： soymilk off-odor volatile substance flavor sensory evaluation

来源：

维普期刊数据库

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

基于用户长短期历史的多兴趣召回算法

引用

南京大学学报（自然科学版） 2024年第1期60卷 12-17页

作者：张旭欧中洪宋美娜北京邮电大学计算机学院北京100876

随着互联网时代的高速发展,用户面临信息过载问题,推荐系统应运而生.推荐系统一般分两个阶段,即推荐召回和推荐排序,推荐召回阶段主要用来筛选出一部分候选集以减小推荐排序阶段的计算压力.多兴趣个性化推荐系统对于每一个用户,算法能... 详细信息

随着互联网时代的高速发展,用户面临信息过载问题,推荐系统应运而生.推荐系统一般分两个阶段,即推荐召回和推荐排序,推荐召回阶段主要用来筛选出一部分候选集以减小推荐排序阶段的计算压力.多兴趣个性化推荐系统对于每一个用户,算法能学习到用户的多种不同的兴趣偏好,然而目前的多兴趣召回算法只考虑了用户短期历史纪录,忽视了用户长期历史纪录中蕴含的丰富信息.针对这一问题,提出一种基于用户长短期历史的多兴趣召回算法,通过不同的神经网络模型结构分别建模用户长短期兴趣偏好,并通过门控融合网络融合用户长短期兴趣偏好,最终得到用户的多个兴趣偏好,实现了个性化推荐召回.在两个公开数据集上的实验证明了模型的有效性.

关键词：推荐系统序列推荐多兴趣长短期历史图神经网络

来源：

维普期刊数据库

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

M^(2)LC-Net: A Multi-Modal Multi-Disease Long-Tailed Classification Network for Real Clinical Scenes

引用

China Communications 2021年第9期18卷 210-220页

作者： Zhonghong Ou Wenjun Chai Lifei Wang Ruru Zhang Jiawen He meina song Lifei Yuan Shengjuan Zhang Yanhui Wang Huan Li Xin Jia Rujian Huang Beijing University of Posts and Telecommunications Beijing 100876China Hebei Eye Hospital Xingtai 054001China

Leveraging deep learning-based techniques to classify diseases has attracted extensive research interest in recent years.Nevertheless,most of the current studies only consider single-modal medical images,and the number of ophthalmic diseases that can be classified is relatively small.Moreover,imbalanced data distribution of different ophthalmic diseases is not taken into consideration,which limits the application of deep learning techniques in realistic clinical scenes.In this paper,we propose a Multimodal Multi-disease Long-tailed Classification Network(M^(2)LC-Net)in response to the challenges mentioned above.M^(2)LC-Net leverages ResNet18-CBAM to extract features from fundus images and Optical Coherence Tomography(OCT)images,respectively,and conduct feature fusion to classify 11 common ophthalmic diseases.Moreover,Class Activation Mapping(CAM)is employed to visualize each mode to improve interpretability of M^(2)LC-Net.We conduct comprehensive experiments on realistic dataset collected from a Grade III Level A ophthalmology hospital in China,including 34,396 images of 11 disease labels.Experimental results demonstrate effectiveness of our proposed model M^(2)LC-Net.Compared with the stateof-the-art,various performance metrics have been improved significantly.Specifically,Cohen’s kappa coefficient κ has been improved by 3.21%,which is a remarkable improvement.

关键词： deep learning multi modal long-tail ophthalmic disease classification

来源：

维普期刊数据库

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

New approach for the digital reconstruction of complex mine faults and its application in mining

引用

International Journal of Coal Science & Technology 2022年第3期9卷 91-105页

作者： Hongwei Wang Zeliang Wang Yaodong Jiang Jiaqi song meina Jia School of Mechanics and Civil Engineering China University of Mining and Technology-BeijingBeijing 100083China State Key Laboratory of Coal Resources and Safe Mining China University of Mining and Technology-BeijingBeijing 100083China

Visualization of complex geological structures can technically support the accurate prediction and prevention of coal mine disasters.This study proposed a new digital reconstruction method to visualize geological structures based on establishing a virtual model in the digital twin system.This methodology for the digital reconstruction of complex fault structures comprises the following four aspects:(1)collection and fdelity of multi-physical feld data of the fault structures,(2)the transmission of multi-physical feld data,(3)the normalization of multi-physical feld data,and(4)digital model reconstruction of fault structures.The key scientifc issues of this methodology to be resolved include in situ fdelity of multi-feld data and normalized programming of multi-source data.In addition,according to the geological background and conditions in Da’anshan coal mine in western Beijing,China,a preliminary attempt is made to reconstruct a digital model of fault and fold structures using the methodology proposed in this study.

关键词： Digital reconstruction Fault structure Multi-physical feld Data collection and fdelity

来源：

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

基于SRIO总线的光纤通信测试系统设计与实现

引用

大气与环境光学学报 2024年第2期19卷 257-264页

作者：赵子豪路美娜许智龙吴洋刘景岩宋茂新骆冬根中国科学院合肥物质科学研究院安徽光学精密机械研究所中国科学院通用光学定标与表征技术重点实验室安徽合肥230031 中国科学技术大学安徽合肥230031

针对基于线阵电荷耦合元件(CCD)成像原理的调焦调平系统强实时性、高同步性及高数据率的特点,设计了一种基于串行RapidIO总线协议的光纤通信测试系统。该系统以现场可编程门阵列(FPGA)为控制核心,以光纤为传输介质,通过串行RapidIO总线... 详细信息

针对基于线阵电荷耦合元件(CCD)成像原理的调焦调平系统强实时性、高同步性及高数据率的特点,设计了一种基于串行RapidIO总线协议的光纤通信测试系统。该系统以现场可编程门阵列(FPGA)为控制核心,以光纤为传输介质,通过串行RapidIO总线协议完成对调焦调平系统性能验证过程中的指令控制和图像数据传输。基于双存储器(FIFO)缓冲的SRIO接口设计使得该测试系统具有较强的通用性和实用性。测试结果表明,该系统光纤接口数据传输速率达440 MB/s,能够满足调焦调平系统性能验证对数据传输速率及正确性的需求。

关键词：调焦调平系统串行RapidIO 现场可编程门阵列光纤通信

来源：

维普期刊数据库博看期刊

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

欢迎您,

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

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

请选择保存的检索档案：

请选择收藏分类：

通借通还

欢迎您,

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

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

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：