检索结果-内蒙古大学图书馆

Multimedia Tools and Applications 2023年 1-2页

作者： K, Vidhya R, Krishna Priya Vidhyalakshmi, M. Ramesh, S. M. L, Bharathi Kshirsagar, Pravin R. Rajaram, A. Tirth, Vineet Department of Computer Science and Engineering KPR Institute of Engineering and Technology Coimbatore India Department of Engineering - Electrical Section University of Technology and Applied Sciences Shinas Campus Sultanate of Oman Department of Electronics and Communication Engineering SRM Institute of Science and Technology Chennai India Department of Signal and Image Processing With Computational Intelligence Saveetha School of Engineering Chennai India Department of Electricals and Electronics Engineering Sathyabama Institute of Science and Technology Chennai India Department of Artificial Intelligence G H Raisoni College of Engineering Nagpur India Department of Electronics and Communication Engineering EGS Pillay Engineering College Nagapattinam India Department of Mechanical Engineering College of Engineering King Khalid University Abha Kingdom of Saudi Arabia

CrackCLF: Automatic Pavement Crack Detection based on Closed-Loop Feedback

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Li, Chong Fan, Zhun Chen, Ying Lin, Huibiao Moretti, Laura Loprencipe, Giuseppe Sheng, Weihua Wang, Kelvin C.P. The Key Lab of Digital Signal and Image Processing of Guangdong Province College of Engineering Shantou University Shantou515063 China The Department of Civil Construction and Environmental Engineering Sapienza University of Rome Rome00184 Italy The School of Electrical and Computer Engineering Oklahoma State University StillwaterOK74078 United States The School of Civil and Environmental Engineering Oklahoma State University StillwaterOK74078 United States

Automatic pavement crack detection is an important task to ensure the functional performances of pavements during their service life. Inspired by deep learning (DL), the encoder-decoder framework is a powerful tool for crack detection. However, these models are usually open-loop (OL) systems that tend to treat thin cracks as the background. Meanwhile, these models can not automatically correct errors in the prediction, nor can it adapt to the changes of the environment to automatically extract and detect thin cracks. To tackle this problem, we embed closed-loop feedback (CLF) into the neural network so that the model could learn to correct errors on its own, based on generative adversarial networks (GAN). The resulting model is called CrackCLF and includes the front and back ends, i.e. segmentation and adversarial network. The front end with U-shape framework is employed to generate crack maps, and the back end with a multi-scale loss function is used to correct higher-order inconsistencies between labels and crack maps (generated by the front end) to address open-loop system issues. Empirical results show that the proposed CrackCLF outperforms others methods on three public datasets. Moreover, the proposed CLF can be defined as a plug and play module, which can be embedded into different neural network models to improve their performances. Copyright © 2023, The Authors. All rights reserved.

关键词： Generative adversarial networks

Monaural Speech Enhancement Using a Multi-Branch Temporal Convolutional Network

学校读者我要写书评

暂无评论

SSRN

SSRN 2022年

作者： Zhang, Qiquan Qian, Xinyuan Nicolson, Aaron Wang, Chenxu Paliwal, Kuldip K. Department of Electrical and Computer Engineering National University of Singapore Singapore Australian e-Health Research Centre CSIRO Australia School of Information and Electrical Engineering Harbin Institute of Technology Weihai264209 China Signal Processing Laboratory Griffith University BrisbaneQLD4160 Australia

Deep learning has achieved substantial improvement on single-channel speech enhancement. The performance of multi-layer perceptrons (MLPs)-based methods is limited by the ability to capture the long-term contextual information. The recurrent neural networks (RNNs), e.g., long short-term memory (LSTM) model, are able to capture the long-term temporal dependencies, but suffer from the high latency and the complex training process. Recently, temporal convolution networks (TCNs), employing 1-D dilated convolution and residual learning, achieves gained considerable success in speech enhancement. In this paper we propose a novel TCN model that employs multi-branch structure, called multi-branch TCN (MB-TCN), for monaural speech enhancement. The MB-TCN exploits split-transform-aggregate design, which is expected to obtain strong representational power with higher parameter efficiency. The MB-TCN model applies 1-D causal dilated CNN in each branch to expand receptive fields for capturing long-term temporal contextual information. Our extensive experimental investigation suggests that MB-TCNs outperform the residual LSTMs (ResLSTMs), TCNs, and the CNNs with dense aggregations (DenseNet) in terms of speech intelligibility and quality, while providing superior parameter efficiency. Furthermore, our proposed MB-TCN model demonstrates highly competitive performance to multiple state-of-the-art speech enhancement methods in terms of five widely used objective metrics. © 2022, The Authors. All rights reserved.

关键词： Convolution

Integrating HL7 FHIR into Clinical Decision Support Systems: A Real-World Application with Pepper Humanoid Robot in Hospital During Doctor Visits

学校读者我要写书评

暂无评论

Integrating HL7 FHIR into Clinical Decision Support Systems:...

International Conference on Mathematics and computers in Sciences and in Industry (MCSI)

作者： Valentino Šafran Samo Horvat Bojan Ilijevec Igor Rober Roj Vojko Flis Izidor Mlakar Laboratory of Digital Signal Processing University of Maribor Faculty of Electrical Engineering and Computer Science Maribor Slovenia Department of Thoracic Surgery University Medical Centre Maribor Maribor Slovenia Clinical Department of Abdominal and General Surgery University Medical Centre Maribor Maribor Slovenia Department of Nurses in Clinic for Surgery University Medical Centre Maribor Maribor Slovenia Clinical Department of Cardiac Surgery University Medical Centre Maribor Maribor Slovenia

ISBN: (数字)9798350377224

ISBN: (纸本)9798350377231

This study explores the integration of HL7 Fast Healthcare Interoperability Resources (FHIR) into a Clinical Decision Support System (CDSS) utilizing a Pepper robot to enhance doctor visits in a hospital setting. We evaluate a real-world implementation where the Pepper robot acts as an interface for data collection from patients, clinical monitors, and hospital information systems. The CDSS leverages FHIR to seamlessly integrate this data and present it visually, aiming to improve clinical decision-making processes for doctors. Our research investigates the potential of this FHIR-based CDSS with a robot interface to optimize healthcare delivery and potentially improve patient outcomes.

关键词： Decision support systems Surveys Hospitals Service robots Soft sensors Transforms User experience Usability Monitoring Interoperability

signal structure information-based data association for maneuvering targets with a convolutional Siamese network

学校读者我要写书评

暂无评论

Signal structure information-based data association for mane...

IEEE International Conference on Radar

作者： Chang Gao Hongwei Liu Pramod K. Varshney Junkun Yan Penghui Wang National Laboratory of Radar Signal Processing Xidian University Xi'an China Department of Electrical Engineering and Computer Science Syracuse University NY USA

ISBN: (纸本)9781665468893

Traditional data association methods usually suffer performance degradation caused by model mismatch when target maneuvers occur. In addition, the signal structure information (SSI) available from the employed signal processing techniques is rarely used in data association due to the lack of appropriate tools. To overcome the model mismatch problem brought by the target maneuvers, we resort to a data driven method to design the data association algorithm. To make full use of the SSI, a convolutional neural network is used to transform the SSI into a feature space, where the similarity of the SSI between two observations can be calculated using the Euclidean distance. The resulting structure is the convolutional Siamese network (CSN). By comparing the similarity with a predetermined margin parameter, the decision about association can be made. Simulation results are presented to show that the use of SSI in the designed CSN can improve the association probability of maneuvering targets.

关键词： Correlation Convolution Simulation Radar Euclidean distance Transforms Radar tracking

Minimal-time Deadbeat Consensus and Individual Disagreement Degree Prediction for High-order Linear Multi-agent Systems

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Hu, Fu-Long Zhang, Hai-Tao Xu, Bowen Hu, Zhe Ren, Wei School of Artificial Intelligence and Automation Key Laboratory of Image Processing and Intelligent Control The State Key Laboratory of Digital Manufacturing Equipment and Technology Huazhong University of Science and Technology Wuhan430074 China School of Artificial Intelligence Optics and Electronics Northwestern Polytechnical University Xi’an710072 China Department of Electrical and Computer Engineering University of California Riverside92521 United States

In this paper, a Hankel matrix-based fully distributed algorithm is proposed to address a minimal-time deadbeat consensus prediction problem for discrete-time high-order multi-agent systems (MASs). Therein, each agent can predict the consensus value with the minimum number of observable historical outputs of its own. Accordingly, compared to most existing algorithms only yielding asymptotic convergence, the present method can attain deadbeat consensus instead. Moreover, based on the consensus value prediction, instant individual disagreement degree value of MASs can be calculated in advance as well. Sufficient conditions are derived to guarantee both the minimal-time deadbeat consensus and the instant individual disagreement degree prediction. Finally, both the effectiveness and superiority of the proposed deadbeat consensus algorithm are substantiated by numerical simulations. Copyright © 2023, The Authors. All rights reserved.

关键词： Multi agent systems

Uncertainty-aware Sampling for Long-tailed Semi-supervised Learning

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Yang, Kuo Li, Duo Hu, Menghan Zhai, Guangtao Yang, Xiaokang Zhang, Xiao-Ping The Shanghai Key Laboratory of Multidimensional Information Processing School of Communication and Electronic Engineering East China Normal University Shanghai200241 China The Kargobot of DiDi Shanghai201210 China The Institute of Image Communication and Network Engineering Shanghai Jiao Tong University Shanghai200240 China Tsinghua Berkeley Shenzhen Institute Shenzhen China The Department of Electrical Computer and Biomedical Engineering Toronto Metropolitan University ONM5B 2K3 Canada

For semi-supervised learning with imbalance classes, the long-tailed distribution of data will increase the model prediction bias toward dominant classes, undermining performance on less frequent classes. Existing methods also face challenges in ensuring the selection of sufficiently reliable pseudo-labels for model training and there is a lack of mechanisms to adjust the selection of more reliable pseudo-labels based on different training stages. To mitigate this issue, we introduce uncertainty into the modeling process for pseudo-label sampling, taking into account that the model performance on the tailed classes varies over different training stages. For example, at the early stage of model training, the limited predictive accuracy of model results in a higher rate of uncertain pseudo-labels. To counter this, we propose an Uncertainty-Aware Dynamic Threshold Selection (UDTS) approach. This approach allows the model to perceive the uncertainty of pseudo-labels at different training stages, thereby adaptively adjusting the selection thresholds for different classes. Compared to other methods such as the baseline method FixMatch, UDTS achieves an increase in accuracy of at least approximately 5.26%, 1.75%, 9.96%, and 1.28% on the natural scene image datasets CIFAR10-LT, CIFAR100-LT, STL-10-LT, and the medical image dataset TissueMNIST, respectively. The source code of UDTS is publicly available at: https://***/yangk/UDTS. Copyright © 2024, The Authors. All rights reserved.

关键词： Uncertainty analysis

Few-Shot Class-Incremental Learning with Prior Knowledge

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Jiang, Wenhao Li, Duo Hu, Menghan Zhai, Guangtao Yang, Xiaokang Zhang, Xiao-Ping Shanghai Key Laboratory of Multidimensional Information Processing School of Communication and Electronic Engineering East China Normal University Shanghai200241 China Kargobot of DiDi Shanghai201210 China Institute of Image Communication and Network Engineering Shanghai Jiao Tong University Shanghai200240 China Tsinghua Berkeley Shenzhen Institute Shenzhen China Department of Electrical Computer and Biomedical Engineering Toronto Metropolitan University ONM5B 2K3 Canada

To tackle the issues of catastrophic forgetting and overfitting in few-shot class-incremental learning (FSCIL), previous work has primarily concentrated on preserving the memory of old knowledge during the incremental phase. The role of pre-trained model in shaping the effectiveness of incremental learning is frequently underestimated in these studies. Therefore, to enhance the generalization ability of the pre-trained model, we propose Learning with Prior Knowledge (LwPK) by introducing nearly free prior knowledge from a few unlabeled data of subsequent incremental classes. We cluster unlabeled incremental class samples to produce pseudo-labels, then jointly train these with labeled base class samples, effectively allocating embedding space for both old and new class data. Experimental results indicate that LwPK effectively enhances the model resilience against catastrophic forgetting, with theoretical analysis based on empirical risk minimization and class distance measurement corroborating its operational principles. The source code of LwPK is publicly available at: https://***/StevenJ308/LwPK. Copyright © 2024, The Authors. All rights reserved.

关键词： Risk assessment

A comparative recognition research on excretory organism in medical applications using neural networks

学校读者我要写书评

暂无评论

Multimedia Tools and Applications 2023年

作者： K, Vidhya R, Krishna Priya Vidhyalakshmi, M. Ramesh, S. M. L, Bharathi Kshirsagar, Pravin R. Rajaram, A. Tirth, Vineet Department of Computer Science and Engineering KPR Institute of Engineering and Technology Tamil Nadu Coimbatore641407 India Department of Engineering - Electrical Section University of Technology and Applied Sciences Shinas Campus Oman Department of Electronics and Communication Engineering SRM Institute of Science and Technology Ramapuram Tamil Nadu Chennai600089 India Department of Signal and Image Processing With Computational Intelligence Saveetha School of Engineering Tamil Nadu Chennai600124 India Department of Electricals and Electronics Engineering Sathyabama Institute of Science and Technology Tamil Nadu Chennai600119 India Department of Artificial Intelligence G H Raisoni College of Engineering Nagpur India Department of Electronics and Communication Engineering EGS Pillay Engineering College Nagapattinam India Department of Mechanical Engineering College of Engineering King Khalid University Asir Abha61421 Saudi Arabia

In the current generation, most individuals are facing many health problems even in the elderly stages and one of the severe touch is caused by to digestive system. Usually, the problems in the digestive system transpire due to improper innards that are present inside it. Even though many experts have described effective solutions to overcome such problems no proper outcomes are achieved. Therefore, the proposed method builds an appropriate device that detects the harmful contents in the entire digestive system which records and stores the data for a continuous period. Additionally, the designed device is incorporated with the neural network using five defined layers where the data is divided into mini segments. This type of data segmentation of input images will reduce the amount of loss in the system thus making the system to be operated at reduced cost. Further equal weights are allocated in the hidden layer using reduced control parameters thereby achieving concentrated attack at several periods. To analyze the effectiveness of neural networks parametric cases are performed and the results prove to be very effective in comparison cases for about 68 percentage. © 2023, The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature.

关键词： Medical applications

A Dynamic Transformer Network for Vehicle Detection

学校读者我要写书评

暂无评论

IEEE Transactions on Consumer Electronics 2025年

作者： Tian, Chunwei Liu, Kai Zhang, Bob Huang, Zhixiang Lin, Chia-Wen Zhang, David Harbin Institute of Technology School of Computer Science and Technology Harbin15001 China Macao Special Administrative Region of China PAMI Research Group University of Macau 999078 China Anhui University Key Laboratory of Intelligent Computing and Signal Processing Ministry of Education Key Laboratory of Electromagnetic Environmental Sensing Hefei230601 China National Tsing Hua University Department of Electrical Engineering Institute of Communications Engineering Hsinchu Taiwan School of Data Science Shenzhen518172 China

Stable consumer electronic systems can assist traffic better. Good traffic consumer electronic systems require collaborative work between traffic algorithms and hardware. However, performance of popular traffic algorithms containing vehicle detection methods based on deep networks via learning data relation rather than learning differences in different lighting and occlusions is limited. In this paper, we present a dynamic Transformer network for vehicle detection (DTNet). DTNet utilizes a dynamic convolution to guide a deep network to dynamically generate weights to enhance adaptability of an obtained detector. Taking into relations of different information account, a mixed attention mechanism based channel attention and Transformer is exploited to strengthen relations of channels and pixels to extract more salient information for vehicle detection. To overcome the drawback of difference in an image account, a translation-variant convolution relies on spatial location information to refine obtained structural information for vehicle detection. Experimental results illustrate that our DTNet is competitive for vehicle detection. Code of the proposed DTNet can be obtained at https://***/hellloxiaotian/DTNet. © 2025 IEEE.

关键词： Dynamic convolution Hybrid Attention Mechanism Transformer Vehicle Detection