检索结果-内蒙古大学图书馆

Proceedings of the 37th International Conference on Neural Information processing Systems

作者： Zhiyuan Zhang Deli Chen Hao Zhou Fandong Meng Jie Zhou Xu Sun National Key Laboratory for Multimedia Information Processing School of Computer Science Peking University and Pattern Recognition Center WeChat AI Tencent Inc. China Pattern Recognition Center WeChat AI Tencent Inc. China National Key Laboratory for Multimedia Information Processing School of Computer Science Peking University

Federated learning algorithms enable neural network models to be trained across multiple decentralized edge devices without sharing private data. However, they are susceptible to backdoor attacks launched by malicious clients. Existing robust federated aggregation algorithms heuristically detect and exclude suspicious clients based on their parameter distances, but they are ineffective on Natural Language processing (NLP) tasks. The main reason is that, although text backdoor patterns are obvious at the underlying dataset level, they are usually hidden at the parameter level, since injecting backdoors into texts with discrete feature space has less impact on the statistics of the model parameters. To settle this issue, we propose to identify backdoor clients by explicitly modeling the data divergence among clients in federated NLP systems. Through theoretical analysis, we derive the f-divergence indicator to estimate the client data divergence with aggregation updates and Hessians. Furthermore, we devise a dataset synthesization method with a Hessian reassignment mechanism guided by the diffusion theory to address the key challenge of inaccessible datasets in calculating clients' data Hessians. We then present the novel Federated F-Divergence-Based Aggregation (Fed-FA) algorithm, which leverages the f-divergence indicator to detect and discard suspicious clients. Extensive empirical results show that Fed-FA outperforms all the parameter distance-based methods in defending against backdoor attacks among various natural language backdoor attack scenarios.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Research on Multi-level Attention-based Human Pose Estimation 2

Research on Multi-level Attention-based Human Pose Estimatio...

引用

2nd International Conference on Algorithms, High Performance Computing and Artificial Intelligence, AHPCAI 2022

作者： Gao, Jun Huang, Hua Li, Qishen Xia, Shaoqi School of Information Engineering Nanchang Hangkong University Jiangxi Nanchang China School of Software Nanchang Hangkong University Jiangxi Nanchang China Key Laboratory of Jiangxi Province for Image Processing and Pattern Recognition Jiangxi Nanchang China

ISBN: (纸本)9798350334784

Detecting human key points from a single image is very challenging due to occlusion, blurring, illumination and scale changes. In this paper, this problem is addressed by designing an effective network structure. Since global and local information plays an important role in reasoning about human body structure and invisible keypoints, Multi-level Attention Network (MAN) is proposed. First, compared with traditional multi-resolution networks, it enables multi-resolution feature maps with greater information variance by generating them directly from the highest resolution feature map, which in turn increases the abundance of feature information after final fusion. Secondly, it effectively integrates global and local information in different resolution feature maps through the Feature Alignment Attention Block(FAAB), and intensifies them in a targeted manner. On the COCO dataset, with HRNet (Sun K. et al [1]) as the baseline network, HRNet of inserted MAN improves 1.1-2.3 AP points over the baseline network. © 2022 IEEE.

关键词： attention mechanism Human Pose Estimation keypoint detection

来源：评论

学校读者我要写书评

暂无评论

A New Deep Fuzzy Based MSER Model for Multiple Document images Classification 3rd

A New Deep Fuzzy Based MSER Model for Multiple Document Imag...

引用

3rd International Conference on pattern recognition and Artificial Intelligence, ICPRAI 2022

作者： Biswas, Kunal Shivakumara, Palaiahnakote Sivanthi, Sittravell Pal, Umapada Lu, Yue Liu, Cheng-Lin Ayub, Mohamad Nizam Bin Computer Vision and Pattern Recognition Unit Indian Statistical Institute Kolkata India Faculty of Computer Science and Information Technology University of Malaya Kuala Lumpur Malaysia Shanghai Key Laboratory of Multidimensional Information Processing East China Normal University Shanghai China National Laboratory of Pattern Recognition Institute of Automation of Chinese Academy of Sciences Beijing China School of Artificial Intelligence University of Chinese Academy Beijing China

ISBN: (纸本)9783031090363

Understanding document images uploaded on social media is challenging because of multiple types like handwritten, printed and scene text images. This study presents a new model called Deep Fuzzy based MSER for classification of multiple document images (like handwritten, printed and scene text). The proposed model detects candidate components that represent dominant information irrespective of the type of document images by combining fuzzy and MSER in a novel way. For every candidate component, the proposed model extracts distance-based features which result in proximity matrix (feature matrix). Further, the deep learning model is proposed for classification by feeding input images and feature matrix as input. To evaluate the proposed model, we create our own dataset and to show effectiveness, the proposed model is tested on standard datasets. The results show that the proposed approach outperforms the existing methods in terms of average classification rate. © 2022, Springer Nature Switzerland AG.

关键词： Image classification

来源：评论

学校读者我要写书评

暂无评论

Research on attention-based multiscale information fusion with the real-time instance segmentation method 2

Research on attention-based multiscale information fusion wi...

引用

2nd International Conference on Algorithms, High Performance Computing and Artificial Intelligence, AHPCAI 2022

作者： He, Yifan Huang, Hua Li, Qishen Zhang, Guiwen School of Information Engineering Nanchang Hangkong University Jiangxi Nanchang China School of Software Nanchang Hangkong University Jiangxi Nanchang China Key Laboratory of Jiangxi Province for Image Processing and Pattern Recognition Jiangxi Nanchang China

ISBN: (纸本)9798350334784

Instance segmentation is a comprehensive computer vision task that involves a wide range of other tasks. Recently, the study of real-time instance segmentation methods has received more attention for the development of autonomous driving. Although existing real-time instance segmentation methods are fast, their accuracy does not meet practical needs. Most methods go for segmentation based on object detection, and their effectiveness is overly dependent on the effectiveness of detection. This paper proposes a new attention-based multiscale information fusion method based on Cheng, T. et al. [1]. Firstly, the PPM module of the baseline network is replaced with the module Multiscale Context Attention (MSCA) designed in this paper based on the baseline network, which uses atrous convolution with different ratios to obtain information of four scales, and then uses non-local attention to enhance the information of features. It can effectively suppress the interference of redundant information on the instance segmentation results. Secondly, a new feature fusion approach is designed, which no longer uses bilinear interpolation, but sub-pixel up sampling combined with attention. We did experiments related to this module on the coco dataset and demonstrated its effectiveness, with a 0.5% improvement over the baseline network. © 2022 IEEE.

关键词： Computer vision

来源：评论

学校读者我要写书评

暂无评论

Generative Zero-shot Image Classification based on Generative Feature Constraint 10

Generative Zero-shot Image Classification based on Generativ...

引用

10th IEEE Joint International Information Technology and Artificial Intelligence Conference, ITAIC 2022

作者： Yu, Xiao Li, Qishen Huang, Hua Li, Qiufeng School of Information Engineering Nanchang Hangkong University Jiangxi Nanchang China School of Software Nanchang Hangkong University Jiangxi Nanchang China Key Laboratory of Jiangxi Province for Image Processing and Pattern Recognition Nanchang China Jiangxi China

ISBN: (数字)9781665422079

ISBN: (纸本)9781665422079

Zero-shot learning (ZSL) utilizes semantic information that is auxiliary information to transfer knowledge from seen classes to unseen classes, thereby realizing the recognition of unseen classes. The generative methods train the model to generate pseudo image features for the unseen classes, thus transforming ZSL into a supervised learning problem. However, the classification performances of these methods still suffer from weak generalization. To address this issue, the paper proposes a novel Generative Adversarial Network based on Generative Feature Constraint (GFCGAN). First, based on the LsrGAN [1], a semantic constraint module is added to restrict the generated image features by keeping the semantic information consistent. Then, in the case of keeping the semantic relationships between classes consistent, the generated image features of unseen classes are farther away from the feature prototypes of their most similar classes in both semantic and feature space so as to train the generative model and generate high-quality pseudo image features. Finally, the trained classifier is used to classify images of unseen classes. Experiments on standard datasets verify the effectiveness of our model and achieve good results. © 2022 IEEE.

关键词： Zero-shot learning

来源：评论

学校读者我要写书评

暂无评论

Multi-scale Features Fusion Network for Single Image Deraining 10

Multi-scale Features Fusion Network for Single Image Deraini...

引用

10th IEEE Joint International Information Technology and Artificial Intelligence Conference, ITAIC 2022

作者： Lai, Yanming Li, Qishen Huang, Hua Li, Qiufeng School of Information Engineering Nanchang Hangkong University Jiangxi Nanchang China School of Software Nanchang Hangkong University Jiangxi Nanchang China Key Laboratory of Jiangxi Province for Image Processing and Pattern Recognition Jiangxi Nanchang China Nanchang Hangkong University Ministry of Education Key Laboratory of Nondestructive Testing Jiangxi China

ISBN: (数字)9781665422079

ISBN: (纸本)9781665422079

Single image rain removal is an important research direction in the field of computer vision. In this paper, the Multi-scale Features Fusion Network (MFFN) is presented for rain removal. MFFN is mainly composed of Multi-features Fusion Module (MFM) and Dual Attention Module (DAM). In the MFM, we make the improved dense block and the dilation convolutions to form the feature extraction branches, which is conducive to improve the receptive fields of network. Subsequently, the bottom-up connection method is adopted between branches to help different branches make full use of image information. Then, feature branches are merged to fusion different features. DAM is consisted of purposed SA Block and standard SE Block. The purpose of DAM is to reduce the non-rain features which is extracted by the network. Experiments show that MFFN can obtain the better result of rain removal than several advanced methods. © 2022 IEEE.

关键词： Computer vision

来源：评论

学校读者我要写书评

暂无评论

Image-to-Class Metric based on Category Traversal for Few-shot Learning 10

Image-to-Class Metric based on Category Traversal for Few-sh...

引用

10th IEEE Joint International Information Technology and Artificial Intelligence Conference, ITAIC 2022

作者： Yuan, Sihao Li, Qishen Huang, Hua Li, Qiufeng School of Information Engineering Nanchang Hangkong University Jiangxi Nanchang China School of Software Nanchang Hangkong University Jiangxi Nanchang China Key Laboratory of Jiangxi Province for Image Processing and Pattern Recognition Jiangxi Nanchang China Jiangxi China

ISBN: (数字)9781665422079

ISBN: (纸本)9781665422079

The purpose of few-shot learning is to learn a classification model with good generalization performance. When there is only one sample or several samples, the model can also be quickly generalized to new categories. In this paper, we use the method based on metric learning to complete the few-shot learning image classification task. In this method, we introduce a feature learning module to improve the representation ability of the feature extraction network. Through category traversal, the module extracts the intra-class common features and inter-class unique features of images. Finally, by comparing the similarity between the input image and the local descriptors of each category, the classification task is completed by the image-to-class measure. This metric is achieved by k-NN search in the local descriptors, in this way, we can make full use of all the local features of a category, thus expressing the distribution of this class more richly and effectively. © 2022 IEEE.

关键词： Image classification

来源：评论

学校读者我要写书评

暂无评论

A Novel Wideband CPW-Fed 5.8GHz RFID Tag Antenna

引用

Chinese Journal of Electronics 2023年第2期21卷 202-208页

作者： Huihui Li Xuanqin Mou Zhen Ji Hang Yu Yan Li Institute of Image Processing and Pattern Recognition Xi'an Jiaotong University Xi'an China College of Computer Science and Software Engineering Shenzhen University Shenzhen China Shenzhen Key Laboratory of Embedded System Design Shenzhen China

A novel wideband 5.8GHz CPW-fed antenna is presented for Radio frequency identification (RFID) tag. Four U-shaped and four L-shaped branches are used as additional resonators to achieve wideband operation. The proposed antenna was analyzed numerically using the Method of moment (MOM) and the Finite element method (FEM). With the antenna size limited to $30\times 30 \text{mm}^{2}$ , the −10dB bandwidth obtained by MOM is 3.235GHz (5.765∼9GHz) and the −9.5dB band-width obtained by FEM is 2.74GHz (5.32∼8.06GHz), corresponding to 55.7% and 47.2% of the center frequency 5.8GHz respectively. Moreover, the simulated results show that the proposed antenna has gain of more than 4.8dBi and the radiation pattern is nearly omnidirectional in the H-plane. The measured −10dB bandwidth is 2.68GHz (5.63GHz∼8.31GHz), 46.2% of the 5.8GHz frequency. Furthermore, there are three measured resonant frequencies at 1.34GHz, 3.23GHz and 5.8GHz with lower than −10dB return loss respectively. The measurement result achieves a wideband RFID tag antenna performance and is in good agreement with the calculated results.

关键词： Antenna measurements Resonant frequency Loss measurement Frequency measurement Finite element analysis Broadband antennas RFID tags

来源：评论

学校读者我要写书评

暂无评论

Decentralized Kernel Ridge Regression Based on Data-Dependent Random Feature

arXiv

引用

arXiv 2024年

作者： Yang, Ruikai He, Fan He, Mingzhen Yang, Jie Huang, Xiaolin The Institute of Image Processing and Pattern Recognition The MOE Key Laboratory of System Control and Information Processing Shanghai Jiao Tong University Shanghai200240 China The STADIUS Center for Dynamical Systems Signal Processing and Data Analytics KU Leuven LeuvenB-3001 Belgium

Random feature (RF) has been widely used for node consistency in decentralized kernel ridge regression (KRR). Currently, the consistency is guaranteed by imposing constraints on coefficients of features, necessitating that the random features on different nodes are identical. However, in many applications, data on different nodes varies significantly on the number or distribution, which calls for adaptive and data-dependent methods that generate different RFs. To tackle the essential difficulty, we propose a new decentralized KRR algorithm that pursues consensus on decision functions, which allows great flexibility and well adapts data on nodes. The convergence is rigorously given and the effectiveness is numerically verified: by capturing the characteristics of the data on each node, while maintaining the same communication costs as other methods, we achieved an average regression accuracy improvement of 25.5% across six real-world data sets. Copyright © 2024, The Authors. All rights reserved.

关键词： Consensus algorithm

来源：评论

学校读者我要写书评

暂无评论

Diffusion Theory as a Scalpel: Detecting and Purifying Poisonous Dimensions in Pre-trained Language Models Caused by Backdoor or Bias

arXiv

引用

arXiv 2023年

作者： Zhang, Zhiyuan Chen, Deli Zhou, Hao Meng, Fandong Zhou, Jie Sun, Xu National Key Laboratory for Multimedia Information Processing School of Computer Science Peking University China Pattern Recognition Center WeChat AI Tencent Inc. China

Pre-trained Language Models (PLMs) may be poisonous with backdoors or bias injected by the suspicious attacker during the fine-tuning process. A core challenge of purifying potentially poisonous PLMs is precisely finding poisonous dimensions. To settle this issue, we propose the Fine-purifying approach, which utilizes the diffusion theory to study the dynamic process of fine-tuning for finding potentially poisonous dimensions. According to the relationship between parameter drifts and Hessians of different dimensions, we can detect poisonous dimensions with abnormal dynamics, purify them by resetting them to clean pre-trained weights, and then fine-tune the purified weights on a small clean dataset. To the best of our knowledge, we are the first to study the dynamics guided by the diffusion theory for safety or defense purposes. Experimental results validate the effectiveness of Fine-purifying even with a small clean dataset. Copyright © 2023, The Authors. All rights reserved.

关键词： Diffusion

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：