检索结果-内蒙古大学图书馆

2022 International Conference on Cloud Computing, Performance Computing, and Deep Learning, CCPCDL 2022

作者： Mao, Chengbo Lu, Menglin Zhao, Tongzhou School of Computer Science and Engineering Wuhan Institute of Technology Wuhan China Hubei Key Laboratory of Intelligent Robot Wuhan Institute of Technology China

ISBN: (纸本)9781510655867

The fine-tune paradigm adopted by traditional paraphrase recognition tasks cannot fully exploit the knowledge of pretrained language models (PLMs). At this stage, the Prompt paradigm reconstructs downstream tasks by constructing templates to make it more suitable for the training form of PLM. However, there are many processes, the code base is not supervised, and a single discrete template limits the model prediction ability. In response to this problem, this paper proposes a method of introducing hybrid prompts based on OpenPrompt. OpenPrompt makes the Prompt process have a unified framework, and hybrid prompts solve the problem that discrete templates cannot fully mine PLM knowledge. This paper first constructs a hybrid template with [mask] slots, and then transforms the original input through the template to obtain xprompt, and then xprompt is input to the bert-base-chinese model of the multi-layer bidirectional transformer based on the attention mechanism for training. When the model trains the optimal prompt, calculate the label with the highest probability of filling in the label set, and finally map the label to the prediction result. The experimental results show that compared with the fine-tune paradigm, the F1 value of the prompt paradigm exceeds 4.69% under the same PLM and hyperparameters. Compared with not using soft prompt, the average accuracy and average F1 value obtained with soft hints are 2.294% and 2.31% higher, respectively. Meanwhile, when the number of soft hints is 6, the accuracy and F1 value reach the highest. © 2022 SPIE.

关键词： Character recognition

来源：评论

学校读者我要写书评

暂无评论

Context-based interactive multimodal sentiment analysis

Context-based interactive multimodal sentiment analysis

引用

2022 International Conference on Cloud Computing, Performance Computing, and Deep Learning, CCPCDL 2022

作者： Ou, Ying Liu, Changhui Liu, Hao School of Computer Science and Engineering Wuhan Institute of Technology Wuhan China Hubei Key Laboratory of Intelligent Robot Wuhan Institute of Technology China

ISBN: (纸本)9781510655867

In view of the singleness of traditional text emotion analysis, multimodal emotion analysis extends emotion analysis to the analysis level of text, image, sound and so on. In order to use the context interaction information expressed in each modality, a multimodal interactive emotion classification model based on video context is proposed in this paper. The ALBERT-BiGRU network structure is built for text feature learning, and the independent BIGRU model is used to extract context features from text, audio and video modality. Based on the attention mechanism, the emotion analysis task is completed after the fusion of the three modal features. Compared with the existing models on MOSI and IEMOCAP data sets, the accuracy and F1 value of emotion classification reached 81.71% and 81.44 on MOSI data set, 66.97% and 67.20 on IEMOCAP data set, which were 1.41% and 2.16% higher than the highest benchmark value respectively, effectively improving the accuracy of multimodal emotion prediction. © 2022 SPIE.

关键词： Sentiment analysis

来源：评论

学校读者我要写书评

暂无评论

Research on Multi-AGVs dynamic scheduling based on deep reinforcement learning 19

Research on Multi-AGVs dynamic scheduling based on deep rein...

引用

19th IEEE International Conference on e-Business Engineering, ICEBE 2023

作者： Qin, Zhuorui Wu, Xiaoqian Zheng, Yimin Wu, Qingyao South China University of Technology School of Software Engineering Guangzhou China Key Laboratory of Big Data and Intelligent Robot Ministry of Education China Shenzhen Bangqi Technology Intelligent Development Co. Ltd Shenzhen China Pazhou Lab Guangzhou China

ISBN: (纸本)9798350325553

Driven by the latest advances in industrial artificial intelligence, automatic guided vehicles (AGVs) have been widely used for material handling in flexible workshops. In the pursuit of agile and flexible production transportation planning, the ability to handle unpredictable and dynamic events is becoming increasingly important. In order to achieve this real-time dynamic control, this study proposes a distributed multi-agent architecture to solve the dynamic Multi-AGVs scheduling problem. Double Deep Q-Network algorithm is used to train the scheduling agent, capture the relationship between production transportation operation information and scheduling goals, and make real-time scheduling decisions for flexible jobs with constant arrival intervals. Specialized state, action, and reward representations are proposed to handle variable specifications of dynamic scheduling problems. The performance of this method under different scene scales was verified through simulation studies. Numerical results show that compared with existing traditional heuristic scheduling strategies, this method not only has superior performance, but also its advantages persist even if the manufacturing system job scale changes. © 2023 IEEE.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Classification of EEG signals based on CNN-Transformer model 20

Classification of EEG signals based on CNN-Transformer model

引用

20th IEEE International Conference on Mechatronics and Automation, ICMA 2023

作者： Liu, Jianwei Dong, Enzeng Tong, Jigang Yang, Sen Du, Shengzhi Tianjin University of Technology Tianjin Key Laboratory for Control Theory & Applications in Complicated Systems Intelligent Robot Laboratory 391 Binshui West Road Tianjin China Tshwane University of Technology Department of Electrical Engineering Pretoria0001 South Africa

ISBN: (纸本)9798350320831

Brain-computer interfaces (BCI) based on EEG have attracted extensive research and attention worldwide, while motor imagery (MI), mental arithmetic (MA), and P300 event-related potentials are a few of the more commonly used *** Transformer(ViT) is a new Transformer model that has superior global processing power compared to Convolutional Neural Networks (CNN) and Recurrent Neural Networks (RNN).In this study, we propose a hybrid CNN-Transformer based model that uses CNN to convolve EEG signals in time and space, followed by ViT for global processing, and finally optimizes the model using 10-run × 10-fold cross-validation and validates it on a publicly available dataset of 29 subjects. Final accuracies of 87.23% and 90.79% were achieved on the MI and MA tasks, respectively. Compared to other literature, the model achieved higher classification accuracies. © 2023 IEEE.

关键词： Brain computer interface

来源：评论

学校读者我要写书评

暂无评论

Fatigue Driving Detection of EEG Signals by LSTM Deep Neural Network with LPSD and DE 20

Fatigue Driving Detection of EEG Signals by LSTM Deep Neural...

引用

20th IEEE International Conference on Mechatronics and Automation, ICMA 2023

作者： Zhang, Hao Dong, Enzeng Tong, Jigang Yang, Sen Du, Shengzhi Tianjin University of Technology Tianjin Key Laboratory for Control Theory & Applications in Complicated Systems Intelligent Robot Laboratory 391 Binshui West Road Tianjin China Tshwane University of Technology Department of Electrical Engineering Pretoria0001 South Africa

ISBN: (纸本)9798350320831

Fatigue driving detection has always been an urgent social problem to be solved, and many complex factors increase the difficulty of detection. Nowadays, more scholars choose to use multi-channel EEG signals for fatigue detection, which is accompanied by the development of deep learning networks. Here, we propose the deep Long Short Term Memory (LSTM) neural network with the Differential Entropy (DE) and Logarithm Power Spectrum Density (LPSD). We first divide the EEG signal into 8-s Hanning windows with 50% overlap, and then use the Short Time Fourier Transform(STFT) on the extracted time window to obtain 50 frequency sub-bands. Finally, LPSD and DE features are extracted from the frequency offspring obtained from the STFT and input into the deep LSTM neural network. The network can choose to learn more task related information by adding soft attention and extract effective time information from EEG signals,thereby improving the model's performance in detecting driving fatigue. In addition, we also collected EEG signals from 8 subjects in both awake and fatigued states,and used the PERCLOS index to judge the degree of fatigue. Using the collected dataset, we compared the effectiveness of the deep LSTM neural network with five other methods. The results indicate that the deep LSTM neural network has a better classification accuracy of 93.12%. It proves that the network has good performance in fatigue detection of EEG signals. © 2023 IEEE.

关键词： Long short-term memory

来源：评论

学校读者我要写书评

暂无评论

A Study of Sentiment Classification Methods with Dual Model Decision Fusion

A Study of Sentiment Classification Methods with Dual Model ...

引用

2022 International Conference on Cloud Computing, Performance Computing, and Deep Learning, CCPCDL 2022

作者： Liu, Hao Liu, Changhui Ou, Ying School of Computer Science and Engineering Wuhan Institute of Technology Wuhan China Hubei Key Laboratory of Intelligent Robot Wuhan Institute of Technology China

ISBN: (纸本)9781510655867

Sentiment recognition is an important part of sentiment computing. To address the problem that traditional Bert models are studied only in textual aspects and a single model has significant uncertainty, an emotion classification recognition method using CM-BERT and Bi-LSTM-DenseNet dual-model decision fusion is proposed. Using both feature-level fusion and decision-level fusion strategies, the superiority of the dual model in multimodal sentiment recognition is highlighted, and the conclusion that the dual model outperforms the single model in multimodal recognition is verified. The model implemented in this paper achieves 64.5% accuracy on a homemade multimodal dataset, a 1% improvement compared to the CM-BERT model, and a 7% improvement compared to the Bi-LSTM-DenseNet model. © 2022 SPIE.

关键词： Long short-term memory

来源：评论

学校读者我要写书评

暂无评论

Modeling and Control Design for Quadrotor Transporting System With a Cable-Suspended Rigid Payload 20

Modeling and Control Design for Quadrotor Transporting Syste...

引用

20th IEEE International Conference on Mechatronics and Automation, ICMA 2023

作者： Xia, Tianqi Dong, Enzeng Yang, Sen Tong, Jigang Du, Shengzhi Tianjin University of Technology Tianjin Key Laboratory for Control Theory & Applications in Complicated Systems Intelligent Robot Laboratory 391 Binshui West Road Tianjin China Tshwane University of Technology Department of Electrical Engineering Pretoria0001 South Africa

ISBN: (纸本)9798350320831

In this paper, facing the aerial transportation task with a large volume cargo, the system of quadrotor with a cable-suspended rigid payload is investigated. The dynamic model of proposed system is built with Lagrange d'Alembert principle. The dynamic model is linearized to design the feedback control method. Finally, two groups of numerical simulation with proposed feedback controller are designed to demonstrate the effectiveness of system model and the results of simulation illustrate satisfactory performance. © 2023 IEEE.

关键词： Dynamic models

来源：评论

学校读者我要写书评

暂无评论

Machine Learning Classification of Riemannian Tangent Spaces Based on MI-BCI 20

Machine Learning Classification of Riemannian Tangent Spaces...

引用

20th IEEE International Conference on Mechatronics and Automation, ICMA 2023

作者： Zhao, Chunli Dong, Enzeng Tong, Jigang Yang, Sen Du, Shengzhi Tianjin University of Technology Tianjin Key Laboratory for Control Theory & Applications in Complicated Systems Intelligent Robot Laboratory 391 Binshui West Road Tianjin China Tshwane University of Technology Department of Electrical Engineering Pretoria0001 South Africa

ISBN: (纸本)9798350320831

Motor imagery brain-computer interfaces (BCIs) have been extensively used in stroke rehabilitation, particularly in China where stroke incidence is high. The application of BCI to control prostheses and robotic arms has shown promising results in improving the quality of life for stroke patients. However, a major challenge in this field is improving the accuracy of motor imagery (MI) classification. Several previous studies have reported that stroke patients exhibit atypical patterns of even-trelated desynchronization (ERD) and event-related synchronization (ERS) activation during motor imagery (MI) tasks. It has been suggested that the brain regions affected by stroke may show reduced excitability and relative insensitivity to external interference compared to those of healthy individuals. Therefore, this study proposes that the classification accuracy of stroke patients in the same MI task would be higher compared to that of healthy volunteers. In this study, we employed a feature extraction algorithm that combines the Filter Bank Common Spatial Pattern (FBCSP) and Riemann geometric mean distance methods to compare the classification accuracy of five different classifiers, with the objective of identifying the most suitable classifier for our analysis. The obtained results revealed that the K-Nearest Neighbors (KNN) classifier yielded the highest accuracy, reaching 87.5% for stroke patients and 85% for healthy volunteers, respectively. In addition to accuracy, we also utilized Kappa values as evaluation metrics, which yielded Kappa results of 0.619 and 0.613 for stroke patients and healthy volunteers, respectively. © 2023 IEEE.

关键词： Brain computer interface

来源：评论

学校读者我要写书评

暂无评论

DGN-MEF: A DenseNet-Based Multi-Exposure Image Fusion Framework using Adaptive Gabor convolution 43

DGN-MEF: A DenseNet-Based Multi-Exposure Image Fusion Framew...

引用

43rd Chinese Control Conference, CCC 2024

作者： Zhang, Mei Jin, Ye Zhu, Jinhui South China University of Technology School of Automation Science and Engineering Guangzhou510641 China South China University of Technology School of Software Engineering Guangzhou510006 China Key Laboratory of Big Data and Intelligent Robot Ministry of Education Guangzhou510006 China

ISBN: (纸本)9789887581581

Multi-exposure image fusion is a cost-effective method to improve the dynamic range of images. For the problems of inadequate detail extraction in bright and dark areas and the over simplicity of existing fusion rules by current deep learning multi-exposure image fusion algorithms, we propose a DenseNet-based framework using adaptive Gabor convolution. Building on the Dense network architecture, we design an adaptive Gabor convolution module, which the parameters of the Gabor filter can be updated along with the network, enhancing the ability of Gabor to extract image features. Additionally, a cross-domain aggregation attention module is proposed, which effectively aggregates and enhances information. Comparing our method with seven advanced methods on the multi-exposure image fusion benchmark dataset, the subjective and objective evaluations outperform the other methods, proving the effectiveness and superiority of our method. The experiments of different exposure images fusion demonstrate the good generalization performance of our method. © 2024 Technical Committee on Control Theory, Chinese Association of Automation.

关键词： Convolution

来源：评论

学校读者我要写书评

暂无评论

Multi-objective optimisation design of cross-shaft based on Kriging response surface optimisation model

引用

International Journal of Wireless and Mobile Computing 2024年第2期26卷 207-214页

作者： Xie, Yuanmin Xiong, Sijie Yun, Juntong Mao, Yanfeng Li, Boao Tang, Xinjie Cao, Yongcheng Key Laboratory of Metallurgical Equipment and Control Technology of Ministry of Education Wuhan University of Science and Technology Hubei Wuhan China Precision Manufacturing Research Institute Wuhan University of Science and Technology Hubei Wuhan China Research Centre for Biomimetic Robot and Intelligent Measurement and Control Wuhan University of Science and Technology Hubei Wuhan China Technology Research and Development Hubei Jingmen Wusan Machinery Equipment Manufacturing Co. Ltd. Hubei Jingmen China

Cross-universal coupling is a key component of the mechanical transmission system and the cross-shaft is the core component of the coupling for torque transmission. Under normal circumstances, cross-shafts are most susceptible to fatigue and deformation, mainly due to the large torques they carry and the irrationality of their structure. Traditional design methods rely on practical experience to determine the key dimensions of the cross-shaft, resulting in long design cycles and low reliability. To address this problem, parametric modelling of the cross-shaft is carried out in this paper and imported into ANSYS Workbench. In addition, static and finite element analyses are carried out to find the weak parts of the cross shaft as the objective function. Finally, sensitivity analysis is carried out using the main structural parameters of the cross-shaft as design variables. Based on the linear correlation matrix and sensitivity graph, the three design variables that have the greatest impact on the objective function, journal height, thickness and body length, are selected. © 2024 Inderscience Enterprises Ltd.

关键词： Sensitivity analysis

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：