检索结果-内蒙古大学图书馆

International Conference on Computer Supported Cooperative Work in Design

作者： Jianhua Gao Jiazheng Man Chunlin Zhao Xiaoming Wu Xiangzhi Liu Huomin Dong Key Laboratory of Computing Power Network and Information Security Ministry of Education Shandong Computer Science Center (National Supercomputer Center in Jinan) Qilu University of Technology (Shandong Academy of Sciences) Jinan China Shandong Provincial Key Laboratory of Industrial Network and Information System Security Shandong Fundamental Research Center for Computer Science Jinan China Shandong Shanke Intelligent Technology Co. Ltd

ISBN: (数字)9798331513054

ISBN: (纸本)9798331513061

As a part of computer vision, object detection is crucial for traffic management, emergency response, autonomous driving vehicles, and smart cities. Despite significant progress in object detection, detecting small objects remains challenging due to their low resolution, which results in little visual information, difficulty in extracting discriminative features, and susceptibility to interference from environmental factors. To address these challenges, we propose SF-DETR, a new model designed specifically for scenarios with small targets. Firstly, We designed a new backbone network that uses partial channel self-attention to replace the backbone network of RTDETR, which can capture both local and global contextual information for extracting and enhancing input features, thereby enhancing the perception of small targets. Secondly, in order to enhance the feature representation ability of the model and better preserve the details of small objects, we proposed Feature Enhancement and Refinement (FER) module, which incorporates a bidirectional fusion mechanism between high-resolution and low-resolution features, allowing for more comprehensive information transfer between features and further improving the effect of multi-scale feature fusion. Finally, we introduce an efficient IoU method (PIoU) which simplifies the computation, speeds up the convergence, and improves the detection accuracy. SF-DETR significantly improves the detection of small targets, outperforming widely used models on various metrics, while significantly reducing model parameters and computational costs compared to RT-DETR. Compared to RTDETR-R34, our model has improved the mAP@50 and mAP@0.5:0.95 by 3.3% and 2.4% respectively on the Visdrone2019 test set.

关键词： Visualization Accuracy Target recognition Smart cities Computational modeling Roads Object detection Feature extraction Computational efficiency Data mining

来源：评论

学校读者我要写书评

暂无评论

Event-aware Video Corpus Moment Retrieval

arXiv

引用

arXiv 2024年

作者： Hou, Danyang Pang, Liang Shen, Huawei Cheng, Xueqi CAS Key Laboratory of AI Security Institute of Computing Technology Chinese Academy of Sciences University of Chinese Academy of Sciences Beijing China CAS Key Lab of Network Data Science and Technology Institute of Computing Technology Chinese Academy of Sciences University of Chinese Academy of Sciences Beijing China

Video Corpus Moment Retrieval (VCMR) is a practical video retrieval task focused on identifying a specific moment within a vast corpus of untrimmed videos using the natural language query. Existing methods for VCMR typically rely on frame-aware video retrieval, calculating similarities between the query and video frames to rank videos based on maximum frame similarity. However, this approach overlooks the semantic structure embedded within the information between frames, namely, the event, a crucial element for human comprehension of videos. Motivated by this, we propose EventFormer, a model that explicitly utilizes events within videos as fundamental units for video retrieval. The model extracts event representations through event reasoning and hierarchical event encoding. The event reasoning module groups consecutive and visually similar frame representations into events, while the hierarchical event encoding encodes information at both the frame and event levels. We also introduce anchor multi-head self-attenion to encourage Transformer to capture the relevance of adjacent content in the video. The training of EventFormer is conducted by two-branch contrastive learning and dual optimization for two sub-tasks of VCMR. Extensive experiments on TVR, ANetCaps, and DiDeMo benchmarks show the effectiveness and efficiency of EventFormer in VCMR, achieving new state-of-the-art results. Additionally, the effectiveness of EventFormer is also validated on partially relevant video retrieval task. © 2024, CC BY.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

PAIR: Complementarity-guided Disentanglement for Composed Image Retrieval

PAIR: Complementarity-guided Disentanglement for Composed Im...

引用

International Conference on Acoustics, Speech, and Signal Processing (ICASSP)

作者： Zhiheng Fu Zixu Li Zhiwei Chen Chunxiao Wang Xuemeng Song Yupeng Hu Liqiang Nie School of Software Shandong University Key Laboratory of Computing Power Network and Information Security Ministry of Education Qilu University of Technology (Shandong Academy of Sciences) School of Computer Science and Technology Shandong University School of Computer Science and Technology Harbin Institute of Technology (Shenzhen)

ISBN: (数字)9798350368741

ISBN: (纸本)9798350368758

Composed Image Retrieval (CIR) is a novel image retrieval paradigm that aims at searching for the target images via the multimodal query including a reference image and a modification text. Although existing works have made significant progress, they overlook the inter-modal coherence and incoherence relations modeling, hindering the retrieval accuracy of CIR models. This limitation is non-trivial due to the following two challenges: 1) inter-modal incoherence and 2) intra-modal entanglement. To address the above challenges, we propose a comPlementArity-guided dIsentanglement network (PAIR), which can disentangle the features of multimodal queries from a semantic coherence perspective, thereby facilitating the identification of both complementary coherent and incoherent features. Furthermore, based on disentangled features, PAIR develops an asymmetric feature composition module, which is designed to enhance the retrieval performance of the model. Extensive experiments on three benchmark datasets demonstrate the superiority of PAIR. The code is available at https://***/***/.

关键词： Codes Accuracy Image retrieval Semantics Coherence Benchmark testing Signal processing Acoustics Speech processing Protection

来源：评论

学校读者我要写书评

暂无评论

CroPrompt: Cross-task Interactive Prompting for Zero-shot Spoken Language Understanding

CroPrompt: Cross-task Interactive Prompting for Zero-shot Sp...

引用

International Conference on Acoustics, Speech, and Signal Processing (ICASSP)

作者： Libo Qin Fuxuan Wei Qiguang Chen Jingxuan Zhou Shijue Huang Jiasheng Si Wenpeng Lu Wanxiang Che School of Computer Science and Engineering Central South University China Key Laboratory of Data Intelligence and Advanced Computing in Provincial Universities Soochow University China Research Center for Social Computing and Information Retrieval Harbin Institute of Technology China Harbin Institute of Technology Shenzhen China Key Laboratory of Computing Power Network and Information Security Ministry of Education Qilu University of Technology (Shandong Academy of Sciences)

ISBN: (数字)9798350368741

ISBN: (纸本)9798350368758

Slot filling and intent detection are two highly correlated tasks in spoken language understanding (SLU). Recent SLU research attempts to explore zero-shot prompting techniques in large language models to alleviate the data scarcity problem. Nevertheless, the existing prompting work ignores the cross-task interaction information for SLU, which leads to sub-optimal performance. To solve this problem, we present the pioneering work of Cross-task Interactive Prompting (CroPrompt) for SLU, which enables the model to interactively leverage the information exchange across the correlated tasks in SLU. Additionally, we further introduce a multi-task self-consistency mechanism to mitigate the error propagation caused by the intent information injection. We conduct extensive experiments on the standard SLU benchmark and the results reveal that CroPrompt consistently outperforms the existing prompting approaches. In addition, the multi-task self-consistency mechanism can effectively ease the error propagation issue, thereby enhancing the performance. We hope this work can inspire more research on cross-task prompting for SLU.

关键词： Large language models Signal processing Benchmark testing Multitasking Filling Acoustics Speech processing Information exchange Standards

来源：评论

学校读者我要写书评

暂无评论

MEDIAN: Adaptive Intermediate-grained Aggregation network for Composed Image Retrieval

MEDIAN: Adaptive Intermediate-grained Aggregation Network fo...

引用

International Conference on Acoustics, Speech, and Signal Processing (ICASSP)

作者： Qinlei Huang Zhiwei Chen Zixu Li Chunxiao Wang Xuemeng Song Yupeng Hu Liqiang Nie School of Software Shandong University Key Laboratory of Computing Power Network and Information Security Ministry of Education Qilu University of Technology (Shandong Academy of Sciences) School of Computer Science and Technology Shandong University School of Computer Science and Technology Harbin Institute of Technology (Shenzhen)

ISBN: (数字)9798350368741

ISBN: (纸本)9798350368758

The Composed Image Retrieval (CIR) task aims to retrieve a target image that meets the requirements based on a given multimodal query (includes a reference image and modification text). Most existing works align multimodal semantics at both local and global granularity. However, they have failed to consider the mining of semantic correspondences at the intermediate-grained level, which has resulted in sub-optimal model performance. In this paper, we propose an adaptive interMEDiate-graIned Aggregation network (MEDIAN). Compared with the conventional CIR models, MEDIAN is capable of generating intermediate-grained feature aggregation supervised signals and constructing graph attention networks to extract intermediate-grained features. Concurrently, MEDIAN also devises cross-modal semantic correspondence aligning guided by the target image, which in turn enables accurate multi-grained feature composition. The superiority of MEDIAN is demonstrated by extensive experiments on three benchmark datasets. Our code is available at https://***/***/.

关键词： Adaptation models Adaptive systems Accuracy Codes Semantics Image retrieval Benchmark testing Signal processing Feature extraction Speech processing

来源：评论

学校读者我要写书评

暂无评论

Reputation based Adaptive Federated Learning in Medical Scenarios

Reputation based Adaptive Federated Learning in Medical Scen...

引用

IEEE International Conference on Data Science in Cyberspace (DSC)

作者： Lianhai Wang Tianrui Liu Yingxiaochun Wang Xinlei Wang Qi Li Key Laboratory of Computing Power Network and Information Security Ministry of Education Shandong Computer Science Center (National Supercomputer Center in Jinan) Qilu University of Technology (Shandong Academy of Sciences) Jinan China Shandong Provincial Key Laboratory of Industrial Network and Information System Security Shandong Fundamental Research Center for Computer Science Jinan China

ISBN: (数字)9798350391367

ISBN: (纸本)9798350391374

Federated Learning (FL) is a distributed machine learning paradigm designed to address data silos and protect data privacy. However, in medical scenarios, the heterogeneity in data quality and the non-independent and identically distributed (Non-IID) nature of medical data from different hospitals can lead to degraded model performance in FL. To tackle this challenge, this paper proposes an adaptive federated learning framework based on reputation. By using a reputation evaluation model, the framework adaptively adjusts the aggregation weights of the models to enhance the performance of the global model. Additionally, our reputation evaluation model, integrated with blockchain technology, can rapidly identify malicious clients and effectively prevent poisoning attacks. Experimental results on two medical datasets demonstrate that this framework achieves higher model accuracy and faster convergence.

关键词： Adaptation models Data privacy Federated learning Hospitals Data integrity Computational modeling Evaluation models Data models security Convergence

来源：评论

学校读者我要写书评

暂无评论

ID-YOLOv7: A Method for Steel Surface Defect Detection Combining Residual Attention and Dynamic Upsampling

ID-YOLOv7: A Method for Steel Surface Defect Detection Combi...

引用

Smart World Congress (SWC), IEEE

作者： Shanshan Liu Xiangzhi Liu Jiazheng Man Bingyang Cao Yunfeng Dong Xiaoming Wu Key Laboratory of Computing Power Network and Information Security Ministry of Education Shandong Computer Science Center (National Supercomputer Center in Jinan) Qilu University of Technology (Shandong Academy of Sciences) Jinan China Shandong Provincial Key Laboratory of Industrial Network and Information System Security Shandong Fundamental Research Center for Computer Science Jinan China

ISBN: (数字)9798331520861

ISBN: (纸本)9798331520878

Industrial defect detection is a crucial challenge in computer vision. The accuracy of detecting defects on steel surfaces is hindered by spatially varying noise. To address this challenge, we propose an efficient method for steel surface defect detection. Firstly, considering the impact of the shape and proportions of bounding boxes on localization accuracy, we developed an Adaptive Shape-Inner Intersection over Union (ASIIoU) loss function. This function dynamically adjusts auxiliary bounding boxes and takes into account their shape and scale to optimize the loss, thereby improving the accuracy of bounding box regression loss. Secondly, to enhance the ability of the model to learn local features, we designed a Content-Guided Residual Attention Block (CGRAB). It captures rich local features across different convolutional layers via skip connections while discarding irrelevant information. Lastly, to avoid the potential issue of detail blurring caused by traditional upsampling methods, we optimized the upsampling process to preserve key details and structural information in the image, reducing unnecessary noise amplification. Experimental validation demonstrates the effectiveness of our method, which achieved a $76.7 \% \mathrm{mAP}$ on the NEU-DET dataset.

关键词： Location awareness Accuracy Shape Noise Focusing Object detection Feature extraction Steel Surface treatment Defect detection

来源：评论

学校读者我要写书评

暂无评论

LCLB: Low Cross-Chain Transaction Ratio and Load Balancing for Blockchain Sharding

LCLB: Low Cross-Chain Transaction Ratio and Load Balancing f...

引用

Robotics, Intelligent Control and Artificial Intelligence (RICAI), International Conference on

作者： Yu Wang Xiaoming Wu Qile Yuan Yunfeng Dong Xiangzhi Liu Key Laboratory of Computing Power Network and Information Security Ministry of Education Shandong Computer Science Center (National Supercomputer Center in Jinan) Qilu University of Technology (Shandong Academy of Sciences) Jinan China Shandong Provincial Key Laboratory of Industrial Network and Information System Security Shandong Fundamental Research Center for Computer Science Jinan China

ISBN: (数字)9798331541699

ISBN: (纸本)9798331541705

As blockchain technology becomes widely applied, handling high volumes of transactions has become a critical challenge. Sharding has emerged as an effective solution to enhance the throughput of blockchain systems. Yet, this approach introduces challenges related to the frequency of cross-shard transactions and workload balancing. High rates of cross-shard transactions can slow overall processing speed and reduce throughput, while uneven load distribution can overburden some shards and underutilize others, impacting overall efficiency. To address these issues, this paper proposes a sharding method named LCLB. This method employs the concept of label propagation to pre-shard accounts with frequent transactions into a designated shard, ensuring a lower ratio of cross-shard transactions. It also uses an elastic load balancing algorithm to dynamically adjust the distribution of accounts, ensuring a relatively balanced processing load across all shards. Our results indicate that LCLB significantly enhances blockchain sharding performance,not only enhancing transaction throughput but also significantly reducing transaction confirmation latency, while ensuring a balanced load across various shards.

关键词： Sharding Heuristic algorithms Throughput Load management Blockchains Partitioning algorithms Delays Resource management Robots Intelligent control

来源：评论

学校读者我要写书评

暂无评论

Research and Application of Speech Transmission Index in Acoustic Leakage Detection

Research and Application of Speech Transmission Index in Aco...

引用

International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI)

作者： Qiuyun Hao FuQiang Wang Peng Zhang XiaoFeng Ma Jianqiang Zhang Wei Zhao Key Laboratory of Computing Power Network and Information Security Ministry of Education Shandong Computer Science Center (National Supercomputer Center in Jinan) Qilu University of Technology (Shandong Academy of Sciences) Jinan China Shandong Provincial Key Laboratory of Industrial Network and Information System Security Shandong Fundamental Research Center for Computer Science Jinan China

ISBN: (数字)9798331507398

ISBN: (纸本)9798331507404

Acoustic information in confidential premise can be leaked through channels such as walls, pipes, doors and windows. Acoustic leakage level needs to be measured through specific indicators. Acoustic leakage evaluation indicator includes subjective evaluation indicator and objective evaluation indicator. The subjective evaluation indicator is a qualitative evaluation of acoustic leakage through subjective auditory perception of human ear, which can directly reflect the degree of acoustic leakage in confidential premise. At present, the objective evaluation indicator is mainly sound transmission loss (STL) to reflect objective degree of acoustic leakage in confidential premise. Therefore, it is relatively singular. Through study of Speech Transmission Index (STI), speech transmission index for public address systems (STIPA), which is a simplified form of FULL STI, is applied to acoustic leakage detection in this paper. STI has been proposed as an objective evaluation indicator and applied to acoustic leakage detection and acoustic information security. In order to confirm the effectiveness of objective evaluation indicator, a large number of related experimental studies have been conducted in the paper. Through experiments, it has been proven that STI can objectively evaluate level of acoustic leakage and level of acoustic information security in confidential premise. STI can better predict level of Chinese clarity under various complex transmission conditions. It provides a reliable basis for speech privacy of confidential premise and evaluation of architectural acoustic, sound reinforcement system and voice communication system.

关键词： Privacy Communication systems Information security Ear Signal processing Propagation losses Acoustics Indexes Reliability Informatics

来源：评论

学校读者我要写书评

暂无评论

A Blockchain-Based Hierarchical Dynamic Collaborative Storage Model

A Blockchain-Based Hierarchical Dynamic Collaborative Storag...

引用

Smart World Congress (SWC), IEEE

作者： Qile Yuan Xiangzhi Liu Yu Wang Yunfeng Dong Haoyu Jia Xiaoming Wu Key Laboratory of Computing Power Network and Information Security Ministry of Education Shandong Computer Science Center (National Supercomputer Center in Jinan) Qilu University of Technology (Shandong Academy of Sciences) Jinan China Shandong Provincial Key Laboratory of Industrial Network and Information System Security Shandong Fundamental Research Center for Computer Science Jinan China

ISBN: (数字)9798331520861

ISBN: (纸本)9798331520878

The pervasive application of blockchain technology has led to a significant increase in data volume, presenting significant challenges to existing storage architectures in environments characterized by large volumes of data and highfrequency transactions. To enhance the scalability of blockchain, this study introduces an innovative framework, Hierarchical Dynamic Storage Management Chain Model, named HDSMC. The model consists of two stages. In the first phase of the model, when multimodal data files are uploaded, LBA techniques are employed to securely eliminate unnecessary redundancy, effectively preventing the storage of duplicate data. In the second phase, further processing is conducted on the generated blocks. This involves analyzing the frequency of block access to compute the variation in block “heat”. Based on the Block Heat (BH), the number of replicas for each block across participating nodes is then determined. Furthermore, the framework implements an adaptive tiering strategy based on data access frequency and its importance, which facilitates the timely migration of data to an appropriate storage level. A simulated Hyperledger Fabric environment was employed to deploy the HDSMC framework and to assess its performance against current storage methodologies. The empirical outcomes indicated that this approach could achieve lower storage expenditures while ensuring high reliability. This research not only presents a novel strategy for blockchain storage management but also suggests potential applications for storage optimization in other big data scenarios.

关键词： Heating systems Time-frequency analysis Sharding Scalability Storage management Redundancy Fabrics Blockchains Reliability Optimization

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：