检索结果-内蒙古大学图书馆

Re-Distributing Facial Features for Engagement Prediction with ModernTCN

Computers, Materials & Continua 2024年第10期81卷 369-391页

作者： Xi Li Weiwei Zhu Qian Li Changhui Hou Yaozong Zhang College of Information and Artificial Intelligence Nanchang Institute of Science and TechnologyNanchang330108China School of Electrical and Information Engineering Wuhan Institute of TechnologyWuhan430205China School of Electronic Information Engineering Wuhan Donghu UniversityWuhan430212China

Automatically detecting learners’engagement levels helps to develop more effective online teaching and assessment programs,allowing teachers to provide timely feedback and make personalized adjustments based on students’needs to enhance teaching *** approaches mainly rely on single-frame multimodal facial spatial information,neglecting temporal emotional and behavioural features,with accuracy affected by significant pose ***,convolutional padding can erode feature maps,affecting feature extraction’s representational *** address these issues,we propose a hybrid neural network architecture,the redistributing facial features and temporal convolutional network(RefEIP).This network consists of three key components:first,utilizing the spatial attention mechanism large kernel attention(LKA)to automatically capture local patches and mitigate the effects of pose variations;second,employing the feature organization and weight distribution(FOWD)module to redistribute feature weights and eliminate the impact of white features and enhancing representation in facial feature ***,we analyse the temporal changes in video frames through the modern temporal convolutional network(ModernTCN)module to detect engagement *** constructed a near-infrared engagement video dataset(NEVD)to better validate the efficiency of the RefEIP *** extensive experiments and in-depth studies,we evaluated these methods on the NEVD and the Database for Affect in Situations of Elicitation(DAiSEE),achieving an accuracy of 90.8%on NEVD and 61.2%on DAiSEE in the fourclass classification task,indicating significant advantages in addressing engagement video analysis problems.

关键词： Engagement prediction spatiotemporal network re-distributing facial features temporal convolutional network

来源：评论

学校读者我要写书评

暂无评论

FIMKD: Feature-Implicit Mapping Knowledge Distillation for RGB-D Indoor Scene Semantic Segmentation

IEEE Transactions on Artificial Intelligence

引用

IEEE Transactions on Artificial Intelligence 2024年第12期5卷 6488-6499页

作者： Zhou, Wujie Xiao, Yuxiang Liu, Yuanyuan Jiang, Qiuping Zhejiang University of Science & Technology School of Information & Electronic Engineering Hangzhou310023 China Nanyang Technological University School of Computer Science and Engineering Singapore308232 Singapore China University of Geosciences School of Computer and Technology Wuhan430074 China Ningbo University School of Information Science and Engineering Ningbo315211 China

Depth images are often used to improve the geometric understanding of scenes owing to their intuitive distance properties. Although there have been significant advancements in semantic segmentation tasks using red-green-blue-depth (RGB-D) images, the complexity of existing methods remains high. Furthermore, the requirement for high-quality depth images increases the model inference time, which limits the practicality of these methods. To address this issue, we propose a feature-implicit mapping knowledge distillation (FIMKD) method and a cross-modal knowledge distillation (KD) architecture to leverage deep modal information for training and reduce the model dependence on this information during inference. The approach comprises two networks: FIMKD-T, a teacher network that uses RGB-D data, and FIMKD-S, a student network that uses only RGB data. FIMKD-T extracts high-frequency information using the depth modality and compensates for the loss of RGB details due to a reduction in resolution during feature extraction by the high-frequency feature enhancement module, thereby enhancing the geometric perception of semantic features. In contrast, the FIMKD-S network does not employ deep learning techniques;instead, it uses a nonlearning approach to extract high-frequency information. To enable the FIMKD-S network to learn deep features, we propose a feature-implicit mapping KD for feature distillation. This mapping technique maps the features in channel and space to a low-dimensional hidden layer, which helps to avoid inefficient single-pattern student learning. We evaluated the proposed FIMKD-S∗ (FIMKD-S with KD) on the NYUv2 and SUN-RGBD datasets. The results demonstrate that both FIMKD-T and FIMKD-S∗ achieve state-of-the-art performance. Furthermore, FIMKD-S∗ provides the best performance balance. © 2020 IEEE.

关键词： Semantic Segmentation

来源：评论

学校读者我要写书评

暂无评论

Belt Sprinkling Detection Method under Complex Working Conditions Based on Machine Vision

IAENG International Journal of Computer Science

引用

IAENG International Journal of Computer science 2025年第3期52卷 546-554页

作者： Wang, Wenzhu Xu, Shaochuan Cong, Bo School of Information Technology Changchun Technical University of Automobile Changchun130013 China School of Electronic and Information Engineering University of Science and Technology Liaoning Anshan114051 China School of Electrical Engineering Changchun Technical University of Automobile Changchun130013 China

Belt conveyors are extensively utilized in material transportation across metallurgy, mining and other industries. Their prolonged operation inevitably leads to issues such as belt deviation and spillage. Currently, belt deviation is primarily due to the uneven distribution of ore, which also contributes to spillage. If these issues are not promptly addressed, they can disrupt production and pose numerous safety risks. The ore transportation process is often characterized by heavy smoke and a complex environment, making manual inspections time-consuming, labor-intensive, and potentially hazardous. This paper introduces a machine vision-based belt sprinkling detection method for complex working conditions. It enhances and processes grayscale images of the belt, collected by a camera, to eliminate smoke interference and accentuate the features of the belt and ore. Edge detection and the Hough transform are employed to pinpoint the belt’s edge, ascertaining the distribution of the belt and ore within it. A GWO-SVM (Grey Wolf Optimizer-Support Vector Machine) model is then constructed to predict the belt’s operational state in real-time, identifying any anomalies to ensure safe production. Experimental comparisons demonstrate that the GWO-SVM model dynamically selects optimal parameters for ‘c’ and ‘g’, yielding accurate classification and detection results. It is characterized by high precision, robust real-time performance, and excellent stability, effectively achieving cost savings and safeguarding production safety. © (2025), (International Association of Engineers). All rights reserved.

关键词： Belt conveyors

来源：评论

学校读者我要写书评

暂无评论

Multi-Link Empowered ISAC Mechanism based on Multiple Radar Subbands Fusion

Multi-Link Empowered ISAC Mechanism based on Multiple Radar ...

引用

2024 International Applied Computational Electromagnetics Society Symposium, ACES-China 2024

作者： Zhang, Ruizhi Zhang, Ying Zhao, Huapeng University of Electronic Science and Technology of China School of Information and Communication Engineering China University of Electronic Science and Technology of China School of Electronic Science and Engineering China

ISBN: (纸本)9798350355581

In this paper, we propose a dual-link cooperative integrated sensing and communication (ISAC) mechanism. Specifically, two transmitters simultaneously send integrated signals across different frequency bands for both sensing and data communication. For radar sensing, the receiver employs multi-band fusion technology to combine the received echo signals into an ultra-wideband (UWB) signal. For communication, the two signals can serve different users by separately modulating the information required by each user. Simulation results demonstrate that the proposed mechanism achieves performance gains in both sensing and communication functions by embedding communication information into specific radar signals. This integration allows for efficient utilization of resources and improved performance in both radar sensing and communication tasks. © 2024 IEEE.

关键词： Ultra-wideband (UWB)

来源：评论

学校读者我要写书评

暂无评论

Design of Belt No-load Detection System Based on Image Processing technology

IAENG International Journal of Computer Science

引用

IAENG International Journal of Computer science 2024年第11期51卷 1731-1739页

作者： Sun, Jiali Xu, Shaochuan Liu, Dehua He, Jing School of Electronic Information University of Science and Technology Liaoning Anshan114051 China School of Control Science and Engineering University of Science and Technology Liaoning Anshan114051 China Anshan Liaoning University of Science and Technology 114051 China

Belts are mainly used for transporting coal. During the conveying process, the belt may be idling. Manual inspections not only lead to reduced efficiency, but also pose safety hazards. As a result, it is proposed a belt no-load monitoring system based on the improved algorithm of support vector machine. The belt is used for on-site monitoring of belt idling problems. An initial image of the belt is acquired by a camera unit and the image is preprocessed. According to the belt images of different working conditions, feature selection is performed for images affected by external light sources, belt surface trace features and ore state distribution. Then support vector machine models using different kernel functions are constructed to classify the belts. Therefore, the belt no-load detection method proposed in this paper can to a certain extent solve the problem that artificial cannot monitor the belt status in real time and timely warning. This method reduces the chance of belt conveyor failures and makes the belt work more efficiently. © (2024), (International Association of Engineers). All rights reserved.

关键词： Belt conveyors

来源：评论

学校读者我要写书评

暂无评论

Tuning-up Learning Parameters for Deep Convolutional Neural Network:A Case Study for Hand-Drawn Sketch Images

引用

Journal of electronic science and technology 2022年第3期20卷 305-318页

作者： Shaukat Hayat Kun She Muhammad Mateen Parinya Suwansrikham Muhammad Abdullah Ahmed Alghaili the School of Information and Software Engineering University of Electronic Science and Technology of ChinaChengdu 610054 the School of Big Data&Software Engineering Chongqing UniversityChongqing 400044 the College of Information Science and Engineering Hunan UniversityChangsha 410082

Several recent successes in deep learning(DL),such as state-of-the-art performance on several image classification benchmarks,have been achieved through the improved ***(HPs)tuning is a key factor affecting the performance of machine learning(ML)*** state-of-the-art DL models use different HPs in different ways for classification tasks on different *** manuscript provides a brief overview of learning parameters and configuration techniques to show the benefits of using a large-scale handdrawn sketch dataset for classification *** analyzed the impact of different learning parameters and toplayer configurations with batch normalization(BN)and dropouts on the performance of the pre-trained visual geometry group 19(VGG-19).The analyzed learning parameters include different learning rates and momentum values of two different optimizers,such as stochastic gradient descent(SGD)and *** analysis demonstrates that using the SGD optimizer and learning parameters,such as small learning rates with high values of momentum,along with both BN and dropouts in top layers,has a good impact on the sketch image classification accuracy.

关键词： Deep learning(DL) hand-drawn sketches learning parameters

来源：评论

学校读者我要写书评

暂无评论

Soft transmission of 3D video for low power and low complexity scenario

引用

Digital Communications and Networks 2023年第3期9卷 769-778页

作者： Ao Li Taihai Yang Wenxin Wu Lei Luo School of Communications and Information Engineering Chongqing University of Posts and TelecommunicationsChongqing400065China School of Computer Science and Technology East China Normal UniversityShanghai200062China School of Information and Communication Engineering University of Electronic Science and Technology of ChinaChengdu611731China

Thanks to the rapid development of naked-eye 3D and wireless communication technology,3D video related applications on mobile devices have attracted a lot of ***,the time-varying characteristics of the wireless channel is very challenging for conventional source-channel coding based transmission ***,the high complexity of source-channel coding based transmission scheme is undesired for low power mobile *** advanced transmission scheme named Softcast was proposed to achieve efficient transmission performance for 2D image/***,it cannot be directly applied to wireless 3D video transmission with high *** paper proposes a more efficient soft transmission scheme for 3D video with a graceful quality adaptation within a wide range of channel Signal-to-Noise Ratio(SNR).The proposed method first extends the linear transform to 4 dimensions with additional view dimension to eliminate the view redundancy,and then metadata optimization and chunk interleaving are designed to further improve the transmission ***,a synthesis distortion based chunk discard strategy is developed to improve the overall 3D video quality under the condition of limited *** experimental results demonstrate that the proposed method significantly improves the 3D video transmission performance over the wireless channel for low power and low complexity scenarios.

关键词： Softcast Low power 3D video Wireless

来源：评论

学校读者我要写书评

暂无评论

ONNXPruner: ONNX-Based General Model Pruning Adapter

引用

IEEE Transactions on Pattern Analysis and Machine Intelligence 2025年第7期47卷 5806-5817页

作者： Ren, Dongdong Li, Wenbin Ding, Tianyu Wang, Lei Fan, Qi Huo, Jing Pan, Hongbing Gao, Yang Nanjing University School of Electronic Science and Engineering China Nanjing University State Key Laboratory for Novel Software Technology School of Intelligence Science and Technology China Microsoft Corporation Applied Sciences Group RedmondWA98052 United States University of Wollongong School of Computing and Information Technology NSW2522 Australia

Recent advancements in model pruning have focused on developing new algorithms and improving upon benchmarks. However, the practical application of these algorithms across various models and platforms remains a significant challenge. To address this challenge, we propose ONNXPruner, a versatile pruning adapter designed for the ONNX format models. ONNXPruner streamlines the adaptation process across diverse deep learning frameworks and hardware platforms. A novel aspect of ONNXPruner is its use of node association trees, which automatically adapt to various model architectures. These trees clarify the structural relationships between nodes, guiding the pruning process, particularly highlighting the impact on interconnected nodes. Furthermore, we introduce a tree-level evaluation method. By leveraging node association trees, this method allows for a comprehensive analysis beyond traditional single-node evaluations, enhancing pruning performance without the need for extra operations. Experiments across multiple models and datasets confirm ONNXPruner's strong adaptability and increased efficacy. Our work aims to advance the practical application of model pruning. © 1979-2012 IEEE.

关键词： Adaptation Models Training Filters Deep Learning Computational Modeling software Interoperability Load Modeling software Algorithms electronic Mail General Model Pruning ONNX Tree Level Evaluation Deep Neural Network Model Pruning Deep Learning Model Architecture Deep Learning Framework Hardware Platform Superior Performance Deep Neural Network Feature Maps Py Torch Image Net Training Images Sparse Matrix Root Node Type Of Operation Single Node Deep Neural Network Model Validation Images Extra Components Importance Of Filtering Pruning Method Squeeze Net

来源：评论

学校读者我要写书评

暂无评论

Named Entity Recognition in electronic Medical Records Incorporating Pre-trained and Multi-Head Attention

IAENG International Journal of Computer Science

引用

IAENG International Journal of Computer science 2024年第4期51卷 401-408页

作者： Yang, Haotian Wang, Li Yang, Yanpeng School of Computer Science and Software Engineering University of Science and Technology Liaoning CO Anshan114051 China College of Computer Science and Software Engineering University of Science and Technology Liaoning CO Anshan114051 China Network Information Centre University of Science and Technology Liaoning CO Anshan114051 China

Chinese Named Entity Recognition (NER) for electronic Medical Records (EMR) is a fundamental task in building a digital hospital and is widely considered to be a sequence annotation problem in the Natural Language Processing domain. (NLP). However, existing deep learning sequence annotation models cannot fully use the large amount of unannotated data for Chinese EMRs that contain a vast number of professional unregistered words, named entities, and inter-of-entity relationships carrying rich professional knowledge. Moreover, the syntactic structure of EMR sentences is complex, and the text is long;the features of the EMR documents often cannot be captured deeply. Aiming at these two problems, this paper proposes a deep learning method that combines Multi-Head Attention with a pre-trained language model. (BERT-BiGRU-Att-CRF). The method uses the BERT pre-trained model to obtain dynamic word vectors combined with contextual information, extracts global semantic features through a Bidirectional Gated Recurrent Unit (BiGRU), obtains augmented semantic features by using the Multi-Head Attention. Finally, using Conditional Random Field (CRF) decoding, outputs the globally optimal label sequence with greatest probability. The model is trained using the CCKS2019 Chinese EMR dataset containing six types of entities: anatomical sites, surgeries, diseases and diagnoses, medicines, laboratory tests, and imaging tests, and good results are achieved with an F1 score of 86.97%. © (2024) International Association of Engineers.

关键词： Random processes

来源：评论

学校读者我要写书评

暂无评论

MSNCIL: A Domain-Agnostic Class-Incremental Learning Method Tailored for Automatic Modulation Recognition

引用

IEEE Communications Letters 2025年第6期29卷 1456-1460页

作者： Deng, Zhiwen Luo, Chunbo Tang, Zixi Luo, Yang University of Electronic Science and Technology of China School of Information and Communication Engineering China

The emergence of new modulation types in 6G challenges the adaptability of deep learning-based automatic modulation recognition (DL-AMR) models. This letter presents multi-state neuron class-incremental learning (MSNCIL), the first domain-agnostic class-incremental learning (CIL) method for AMR. Leveraging the sparsity of wireless signal features, MSNCIL dynamically partitions a DL-AMR model into specialized sub-models, each dedicated to different modulation types. In each session, neurons are selected based on activation values, trained, frozen, and assigned state values. During inference, the session ID of a test sample is identified, which directs the corresponding neurons for recognition. Extensive experiments confirm MSNCIL's effectiveness. © 1997-2012 IEEE.

关键词： Deep learning

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：