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： Shangshang Yang Xiaoshan Yu Ye Tian Xueming Yan Haiping Ma Xingyi Zhang Anhui University and Key Laboratory of Intelligent Computing and Signal Processing of Ministry of Education Guangdong University of Foreign Studies Anhui University and Key Laboratory of Intelligent Computing and Signal Processing of Ministry of Education and Department of Information Materials and Intelligent Sensing Laboratory of Anhui Province

Transformer has achieved excellent performance in the knowledge tracing (KT) task, but they are criticized for the manually selected input features for fusion and the defect of single global context modelling to directly capture students' forgetting behavior in KT, when the related records are distant from the current record in terms of time. To address the issues, this paper first considers adding convolution operations to the Transformer to enhance its local context modelling ability used for students' forgetting behavior, then proposes an evolutionary neural architecture search approach to automate the input feature selection and automatically determine where to apply which operation for achieving the balancing of the local/global context modelling. In the search space design, the original global path containing the attention module in Transformer is replaced with the sum of a global path and a local path that could contain different convolutions, and the selection of input features is also considered. To search the best architecture, we employ an effective evolutionary algorithm to explore the search space and also suggest a search space reduction strategy to accelerate the convergence of the algorithm. Experimental results on the two largest and most challenging education datasets demonstrate the effectiveness of the architecture found by the proposed approach.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Few-Shot Object Detection Based on the Transformer and High-Resolution Network

引用

computers, Materials & Continua 2023年第2期74卷 3439-3454页

作者： Dengyong Zhang Huaijian Pu Feng Li Xiangling Ding Victor S.Sheng Hunan Provincial Key Laboratory of Intelligent Processing of Big Data on Transportation Changsha University of Science and TechnologyChangsha410114China School of Computer and Communication Engineering Changsha University of Science and TechnologyChangsha410114China School of Computer Science and Engineering Hunan University of Science and TechnologyXiangtan411004China Department of Computer Science Texas Tech UniversityLubbock79409TXUSA

Now object detection based on deep learning tries different *** uses fewer data training networks to achieve the effect of large dataset ***,the existing methods usually do not achieve the balance between network parameters and training *** makes the information provided by a small amount of picture data insufficient to optimize model parameters,resulting in unsatisfactory detection *** improve the accuracy of few shot object detection,this paper proposes a network based on the transformer and high-resolution feature extraction(THR).High-resolution feature extractionmaintains the resolution representation of the *** and spatial attention are used to make the network focus on features that are more useful to the *** addition,the recently popular transformer is used to fuse the features of the existing *** compensates for the previous network failure by making full use of existing object *** on the Pascal VOC and MS-COCO datasets prove that the THR network has achieved better results than previous mainstream few shot object detection.

关键词： Object detection few shot object detection transformer high-resolution

来源：评论

学校读者我要写书评

暂无评论

EAFormer: Scene Text Segmentation with Edge-Aware Transformers

arXiv

引用

arXiv 2024年

作者： Yu, Haiyang Fu, Teng Li, Bin Xue, Xiangyang Shanghai Key Laboratory of Intelligent Information Processing School of Computer Science Fudan University China

Scene text segmentation aims at cropping texts from scene images, which is usually used to help generative models edit or remove texts. The existing text segmentation methods tend to involve various text-related supervisions for better performance. However, most of them ignore the importance of text edges, which are significant for downstream applications. In this paper, we propose Edge-Aware Transformers, termed EAFormer, to segment texts more accurately, especially at the edge of texts. Specifically, we first design a text edge extractor to detect edges and filter out edges of non-text areas. Then, we propose an edge-guided encoder to make the model focus more on text edges. Finally, an MLP-based decoder is employed to predict text masks. We have conducted extensive experiments on commonly-used benchmarks to verify the effectiveness of EAFormer. The experimental results demonstrate that the proposed method can perform better than previous methods, especially on the segmentation of text edges. Considering that the annotations of several benchmarks (e.g., COCO_TS and MLT_S) are not accurate enough to fairly evaluate our methods, we have relabeled these datasets. Through experiments, we observe that our method can achieve a higher performance improvement when more accurate annotations are used for training. The code and datasets are available at https://***/EAFormer/. © 2024, CC BY.

关键词： Image segmentation

来源：评论

学校读者我要写书评

暂无评论

E2E-MFERC:AMulti-Face Expression Recognition Model for Group Emotion Assessment

引用

computers, Materials & Continua 2024年第4期79卷 1105-1135页

作者： Lin Wang Juan Zhao Hu Song Xiaolong Xu Jiangsu Key Laboratory of Big Data Security&Intelligent Processing Nanjing University of Posts and TelecommunicationsNanjing210042China School of Network Security Jinling Institute of TechnologyNanjing211169China State Grid Jiangsu Electric Power Company Limited Nanjing210000China School of Computer Science Nanjing University of Posts and TelecommunicationsNanjing210042China

In smart classrooms, conducting multi-face expression recognition based on existing hardware devices to assessstudents’ group emotions can provide educators with more comprehensive and intuitive classroom effect analysis,thereby continuouslypromotingthe improvementof teaching ***,most existingmulti-face expressionrecognition methods adopt a multi-stage approach, with an overall complex process, poor real-time performance,and insufficient generalization ability. In addition, the existing facial expression datasets are mostly single faceimages, which are of low quality and lack specificity, also restricting the development of this research. This paperaims to propose an end-to-end high-performance multi-face expression recognition algorithm model suitable forsmart classrooms, construct a high-quality multi-face expression dataset to support algorithm research, and applythe model to group emotion assessment to expand its application value. To this end, we propose an end-to-endmulti-face expression recognition algorithm model for smart classrooms (E2E-MFERC). In order to provide highqualityand highly targeted data support for model research, we constructed a multi-face expression dataset inreal classrooms (MFED), containing 2,385 images and a total of 18,712 expression labels, collected from smartclassrooms. In constructing E2E-MFERC, by introducing Re-parameterization visual geometry group (RepVGG)block and symmetric positive definite convolution (SPD-Conv) modules to enhance representational capability;combined with the cross stage partial network fusion module optimized by attention mechanism (C2f_Attention),it strengthens the ability to extract key information;adopts asymptotic feature pyramid network (AFPN) featurefusion tailored to classroomscenes and optimizes the head prediction output size;achieves high-performance endto-end multi-face expression detection. Finally, we apply the model to smart classroom group emotion assessmentand provide design refe

关键词： Multi-face expression recognition smart classroom end-to-end detection group emotion assessment

来源：评论

学校读者我要写书评

暂无评论

Improving Viewpoint-Independent Object-Centric Representations through Active Viewpoint Selection

arXiv

引用

arXiv 2024年

作者： Huang, Yinxuan Gao, Chengmin Li, Bin Xue, Xiangyang Shanghai Key Laboratory of Intelligent Information Processing School of Computer Science Fudan University China

Given the complexities inherent in visual scenes, such as object occlusion, a comprehensive understanding often requires observation from multiple viewpoints. Existing multi-viewpoint object-centric learning methods typically employ random or sequential viewpoint selection strategies. While applicable across various scenes, these strategies may not always be ideal, as certain scenes could benefit more from specific viewpoints. To address this limitation, we propose a novel active viewpoint selection strategy. This strategy predicts images from unknown viewpoints based on information from observation images for each scene. It then compares the object-centric representations extracted from both viewpoints and selects the unknown viewpoint with the largest disparity, indicating the greatest gain in information, as the next observation viewpoint. Through experiments on various datasets, we demonstrate the effectiveness of our active viewpoint selection strategy, significantly enhancing segmentation and reconstruction performance compared to random viewpoint selection. Moreover, our method can accurately predict images from unknown viewpoints. Copyright © 2024, The Authors. All rights reserved.

关键词： Active learning

来源：评论

学校读者我要写书评

暂无评论

DiffTimb: Diffusion Models for Many-to-Many Timbre Transfer 10th

DiffTimb: Diffusion Models for Many-to-Many Timbre Transfer

引用

10th Conference on Sound and Music Technology, CSMT 2023

作者： Gan, Xu Wu, Yifei Liu, Xinlu Duan, Ruilei Li, Wei School of Computer Science and Technology Fudan University Shanghai200438 China Department of Music Engineering ZheJiang Conservatory of Music Hangzhou310024 China Shanghai Key Laboratory of Intelligent Information Processing Fudan University Shanghai200433 China

ISBN: (纸本)9789819779611

The goal of the timbre transfer task is to apply the auditory characteristics of one instrument to another instrument. Past timbre transfer methods faced numerous challenges due to the limitations of the selected generation model. These challenges include difficulties in training, ambiguous results, poor diversity, and slow generation speed. To address these issues, we propose a diffusion model for timbre transfer, DiffTimb, which can be combined with vocoder to achieve single-note timbre transfer. Unlike traditional models that require training a specific model for each pair of timbre transfers, DiffTimb enables the conversion of various musical instrument timbres received in a single model, namely, many-to-many timbre transfer. We compare two Unet-based denoising frameworks and propose two feature fusion methods to fuse timbre and pitch features. We conduct unconditional and conditional generation pre-experiments to verify the generation ability of DiffTimb. To improve the generation effect, we clean the Nsynth dataset and get two subsets for our task. Experimental results show that DiffTimb has outperformed previous methods in both subjective and objective evaluations. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Audio acoustics

来源：评论

学校读者我要写书评

暂无评论

ASTF: Visual Abstractions of Time-Varying Patterns in Radio signals

引用

IEEE Transactions on Visualization and computer Graphics 2023年第1期29卷 214-224页

作者： Zhao, Ying Ge, Luhao Xie, Huixuan Bai, Genghuai Zhang, Zhao Wei, Qiang Lin, Yun Liu, Yuchao Zhou, Fangfang School of Computer Science and Engineering Central South University Changsha China National Key Laboratory of Science and Technology on Blind Signal Processing Chengdu China College of Information and Communication Engineering Harbin Engineering University Harbin China China Research Institute of Radiowave Propagation Qingdao China

A time-frequency diagram is a commonly used visualization for observing the time-frequency distribution of radio signals and analyzing their time-varying patterns of communication states in radio monitoring and management. While it excels when performing short-term signal analyses, it becomes inadaptable for long-term signal analyses because it cannot adequately depict signal time-varying patterns in a large time span on a space-limited screen. This research thus presents an abstract signal time-frequency (ASTF) diagram to address this problem. In the diagram design, a visual abstraction method is proposed to visually encode signal communication state changes in time slices. A time segmentation algorithm is proposed to divide a large time span into time slices. Three new quantified metrics and a loss function are defined to ensure the preservation of important time-varying information in the time segmentation. An algorithm performance experiment and a user study are conducted to evaluate the effectiveness of the diagram for long-term signal analyses. © 2022 IEEE.

关键词： Data visualization

来源：评论

学校读者我要写书评

暂无评论

AdvQDet: Detecting Query-Based Adversarial Attacks with Adversarial Contrastive Prompt Tuning

arXiv

引用

arXiv 2024年

作者： Wang, Xin Chen, Kai Ma, Xingjun Chen, Zhineng Chen, Jingjing Jiang, Yu-Gang Shanghai Key Laboratory of Intelligent Information Processing School of Computer Science Fudan University China

Deep neural networks (DNNs) are known to be vulnerable to adversarial attacks even under a black-box setting where the adversary can only query the model. Particularly, query-based blackbox adversarial attacks estimate adversarial gradients based on the returned probability vectors of the target model for a sequence of queries. During this process, the queries made to the target model are intermediate adversarial examples crafted at the previous attack step, which share high similarities in the pixel space. Motivated by this observation, stateful detection methods have been proposed to detect and reject query-based attacks. While demonstrating promising results, these methods either have been evaded by more advanced attacks or suffer from low efficiency in terms of the number of shots (queries) required to detect different attacks. Arguably, the key challenge here is to assign high similarity scores for any two intermediate adversarial examples perturbed from the same clean image. To address this challenge, we propose a novel Adversarial Contrastive Prompt Tuning (ACPT) method to robustly fine-tune the CLIP image encoder to extract similar embeddings for any two intermediate adversarial queries. With ACPT, we further introduce a detection framework AdvQDet that can detect 7 state-of-the-art query-based attacks with > 99% detection rate within 5 shots. We also show that ACPT is robust to 3 types of adaptive attacks. Code is available at https://***/xinwong/AdvQDet. Copyright © 2024, The Authors. All rights reserved.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

Generalized Self-Adaption Network for Domain Adaptation 2

Generalized Self-Adaption Network for Domain Adaptation

引用

2nd International Conference on Artificial Intelligence and intelligent Information processing, AIIIP 2023

作者： Bian, Yongheng Ke, Xiao College of Computer and Data Science Fuzhou University Fujian Provincial Key Laboratory of Networking Computing and Intelligent Information Processing Fuzhou350116 China Key Laboratory of Spatial Data Mining & Information Sharing Ministry of Education Fuzhou350003 China

ISBN: (纸本)9798350371451

The purpose of domain adaptation is to transfer knowledge learned in the labeled source domain to unlabeled but related target domains without requiring a large number of target domain labels. The latest method of domain adaptation mainly adjusts the distribution of the source and target domains to adapt to the target domain. Despite significant success, when there are significant distribution differences between the source and target domains, the adaptability in the target domain still decreases, which may be due to insufficient exploration of domain specific features leading to domain offset issues. To address this issue, we propose the GSA-NET framework in this article. We designed an adversarial dropout method to improve the generalization performance of the model in the target domain by pushing the decision boundary away from the feature dense region. At the same time, in order to effectively match the distribution of advanced features across domains, we deploy feature adaptation modules before the classification layer, which can significantly reduce domain differences. Our experiments on two cross domain benchmarks, Office Home and Office-31, have shown that the proposed method is significantly superior to existing methods. © 2023 IEEE.

关键词： computer vision

来源：评论

学校读者我要写书评

暂无评论

Multi-view Fusion Enhanced Social Text Representation for Depression Detection 11th

Multi-view Fusion Enhanced Social Text Representation for D...

引用

11th International Joint Conference on Rough Sets, IJCRS 2025

作者： Chen, Jie You, Xuetong Zhao, Shu Xiao, Jingxuan Zhang, Yanping Key Laboratory of Intelligent Computing and Signal Processing Ministry of Education Anhui Province Hefei230601 China School of Computer Science and Technology Anhui University Anhui Province Hefei230601 China Information Materials and Intelligent Sensing Laboratory of Anhui Province Anhui Province Hefei230601 China Faculty of Innovation Engineering Department of Computer Science Macau University of Science and Technology Macau China

ISBN: (纸本)9783031927409

Depression has become a major global public health challenge. Social media texts reflect users’ emotional and psychological states, providing a valuable data source for depression detection through natural language analysis. However, current research still faces limitations in modeling the collaborative representation of user texts and dynamic behaviors. To address this, we propose MVDep, a model that integrates multi-view features to enhance social text representation for depression detection. Specifically, we enhance text representation by investigating depression from multi-view, including social media texts, user profiles, and social behaviors, enabling more effective depression detection. First, we conduct initial filtering of social media texts using the Diagnostic and Statistical Manual of Mental Disorders (DSM-5), and then screen out the uncertain domain of data samples based on the three-way decision (3WD). Next, we extract quantified user profiles and social behavioral features to enhance the representation of depression tendencies for uncertain domain samples. Finally, we fuse the multi-view features derived from social media texts, user profiles and social behaviors, improving the model’s ability to detect depression. We evaluate our model on the SWDD dataset, demonstrating its superiority over state-of-the-art baseline models in depression detection tasks. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Tweets

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：