检索结果-内蒙古大学图书馆

Improving Incremental Learning: A Closer Look at the Softmax Function

学校读者我要写书评

暂无评论

SSRN

SSRN 2024年

作者： Zhai, Zheng Zhang, Jiali Wang, Haiyu Wu, Mingxin Yang, Keshun Qiao, Xiaoyan Sun, Qiang Beijing Normal University No.18 Jinfeng Road Guangdong Zhuhai519087 China Shandong Technology and Business University Shandong Yantai China Yantai Key Laboratory of Big Data Modeling and Intelligent Computing Shandong China Immersion Technology and Evaluation Shandong Engineering Research Center Shandong China School of Mathematics Sichuan University Chengdu China College of Liberal Arts and Sciences University of Illinois Urbana-Champaign IL United States Department of Statistical Sciences University of Toronto ON Canada Department of Computer Science University of Toronto ON Canada Department of Statistics and Data Science MBZUAI Abu Dhabi United Arab Emirates

This paper investigates the limitations of the widely adopted softmax cross-entropy loss in incremental learning problems. Specifically, we highlight how the shift-invariant property of this loss function can lead to multiple optimal solutions and imbalanced weights across different tasks, which exacerbates catastrophic forgetting. To address these issues, we propose two key modifications to the objective function: (1) regularizing or replacing the original joint distillation loss with a summation of task-specific distillation losses, and/or (2) utilizing a shift-sensitive prediction loss function, such as the binary cross-entropy loss. Numerical experiments validate the effectiveness of these proposals, demonstrating improved prediction accuracy and a significant reduction in forgetting. © 2024, The Authors. All rights reserved.

关键词： Adversarial machine learning

A Vision-Language Foundation Model to Enhance Efficiency of Chest X-ray Interpretation

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Chen, Zhihong Varma, Maya Xu, Justin Paschali, Magdalini Van Veen, Dave Johnston, Andrew Youssef, Alaa Blankemeier, Louis Bluethgen, Christian Altmayer, Stephan Valanarasu, Jeya Maria Jose Muneer, Mohamed Siddig Eltayeb Reis, Eduardo Pontes Cohen, Joseph Paul Olsen, Cameron Abraham, Tanishq Mathew Tsai, Emily B. Beaulieu, Christopher F. Jitsev, Jenia Gatidis, Sergios Delbrouck, Jean-Benoit Chaudhari, Akshay S. Langlotz, Curtis P. Stanford Center for Artificial Intelligence in Medicine and Imaging Stanford University Palo AltoCA United States Department of Radiology Stanford University StanfordCA United States Department of Computer Science Stanford University StanfordCA United States Big Data Institute University of Oxford Oxford United Kingdom Department of Electrical Engineering Stanford University StanfordCA United States Department of Radiology University Hospital Zurich Zürich Switzerland Stability AI London United Kingdom Jülich Supercomputing Centre Jülich Germany LAION Germany Department of Biomedical Data Science Stanford University StanfordCA United States Department of Medicine Stanford University StanfordCA United States

Over 1.4 billion chest X-rays (CXRs) are performed annually due to their cost-effectiveness as an initial diagnostic test. This scale of radiological studies provides a significant opportunity to streamline CXR interpretation and documentation. While foundation models are a promising solution, the lack of publicly available large-scale datasets and benchmarks inhibits their iterative development and real-world evaluation. To overcome these challenges, we constructed a large-scale dataset (CheXinstruct), which we utilized to train a vision-language foundation model (CheXagent). We systematically demonstrated competitive performance across eight distinct task types on our novel evaluation benchmark (CheXbench). Beyond technical validation, we assessed the real-world utility of CheXagent in directly drafting radiology reports. Our clinical assessment with eight radiologists revealed a 36% time saving for residents using CheXagent-drafted reports, while attending radiologists showed no significant time difference editing resident-drafted or CheXagent-drafted reports. The CheXagent-drafted reports improved the writing efficiency of both radiology residents and attending radiologists in 81% and 61% of cases, respectively, without loss of quality. Overall, we demonstrate that CheXagent can effectively perform a variety of CXR interpretation tasks and holds potential to assist radiologists in routine clinical workflows. © 2024, CC BY.

关键词： Visual languages

Transformer-based radio modulation mode recognition 4

学校读者我要写书评

暂无评论

Transformer-based radio modulation mode recognition

2022 4th International Conference on Wireless Communications and Smart Grid, ICWCSG 2022

作者： Li, Lingyun Qin, Chanchan Li, Guoqing Hu, Shengbo Xie, Yike Lei, Zhenwei School of Big Data and Computer Science Guizhou Normal University The University Town Guian New Area Guizhou Guiyang550025 China Center for RFID and WSN Engineering Department of Education Guizhou Normal University The University Town Guian New Area Guizhou Guiyang550025 China College of Physical Science and Technology Central China Normal University No.152 Luoyu Road Hubei Wuhan430079 China College of Physical Science and Technology Central China Normal University NO.152 Luoyu Road Hubei Wuhan430079 China

The wireless communication technology develop rapidly, signal modulation mode technology becomes increasingly important. Therefore, in order to enhance the ability of radio signal modulation mode recognition in the complex electromagnetic environment, a novel radio signal modulation mode recognition method is proposed on account of Transformer network. Firstly, raw data were split with a fixed size window. Then, the segmented data was projected into a vector sequence and fed into the Transformer module to model and mine the relationship between signal waveform and modulation method to achieve identification. Extensive experiments were conducted on the RML2016.10a data set generated by GNU radio in the real system. The experimental results show that the proposed Transformer-B/8×2 achieves optimal results for modulation mode recognition with signal-to-noise from -20dB to 18dB, and recognition accuracy reaches 86.75%, verifying the practicability and validity of the forenamed algorithm. © Published under licence by IOP Publishing Ltd.

关键词： Signal to noise ratio

Diagnostic accuracy of pre-operative breast magnetic resonance imaging (MRI) in predicting axillary lymph node metastasis: variations in intrinsic subtypes, and strategy to improve negative predictive value—an analysis of 2473 invasive breast cancer patients

学校读者我要写书评

暂无评论

Breast Cancer 2023年第6期30卷 976-985页

作者： Chen, Shu-Tian Lai, Hung-Wen Chang, Julia Huei-Mei Liao, Chiung-Ying Wen, Tzu-Cheng Wu, Wen-Pei Wu, Hwa-Koon Lin, Ying-Jen Chang, Yu-Jun Chen, Shou-Tung Chen, Dar-Ren Huang, Hsin-I Hung, Che-Lun Department of Diagnostic Radiology Chang Gung Memorial Hospital - Chiayi Branch Chiayi Taiwan Institute of Biomedical Informatics National Yang Ming Chiao Tung University No.155 Sec. 2 Linong St. Beitou Dist. Taipei 11221 Taiwan Department of Biomedical Imaging and Radiological Sciences National Yang Ming Chiao Tung University Taipei Taiwan School of Medicine National Yang Ming Chiao Tung University Taipei Taiwan Endoscopy and Oncoplastic Breast Surgery Center Changhua Christian Hospital 135 Nanxiao Street Changhua 500 Taiwan Division of General Surgery Changhua Christian Hospital Changhua Taiwan Comprehensive Breast Cancer Center Changhua Christian Hospital Changhua Taiwan Tumor Center Changhua Christian Hospital Changhua Taiwan Department of Radiology Changhua Christian Hospital Changhua Taiwan Big Data Center Changhua Christian Hospital Changhua Taiwan Kaohsiung Medical University Kaohsiung Taiwan Division of Breast Surgery Yuanlin Christian Hospital Yuanlin Taiwan School of Medicine Chung Shan Medical University Taichung Taiwan Department of Pathology Changhua Christian Hospital Changhua Taiwan Department of Information Management National Sun Yat-Sen University Kaohsiung Taiwan We-Sing Breast Hospital Kaohsiung Taiwan Department of Computer Science and Communication Engineering Providence University Taichung Taiwan

Background: The value and utility of axillary lymph node (ALN) evaluation with MRI in breast cancer were not clear for various intrinsic subtypes. The aim of the current study is to test the potential of combining breast MRI and clinicopathologic factors to identify low-risk groups of ALN metastasis and improve diagnostic performance. Material and methods: Patients with primary operable invasive breast cancer with pre-operative breast MRI and post-operative pathologic reports were retrospectively collected from January 2009 to December 2021 in a single institute. The concordance of MRI and pathology of ALN status were determined, and also analyzed in different intrinsic subtypes. A stepwise strategy was designed to improve MRI-negative predictive value (NPV) on ALN metastasis. Results: 2473 patients were enrolled. The diagnostic performance of MRI in detecting metastatic ALN was significantly different between intrinsic subtypes (p = 0.007). Multivariate analysis identified tumor size and histologic type as independent predictive factors of ALN metastases. Patients with HER-2 (MRI tumor size ≤ 2 cm), or TNBC (MRI tumor size ≤ 2 cm) were found to have MRI–ALN-NPV higher than 90%, and these false cases were limited to low axillary tumor burden. Conclusion: The diagnostic performance of MRI to predict ALN metastasis varied according to the intrinsic subtype. Combined pre-operative clinicopathologic factors and intrinsic subtypes may increase ALN MRI NPV, and further identify some groups of patients with low risks of ALN metastasis, high NPV, and low burdens of axillary disease even in false-negative cases. © 2023, The Author(s).

关键词： Breast magnetic resonance imaging (MRI) Intrinsic subtype Lymph node metastasis Negative predictive value (NPV) Sentinel lymph node biopsy (SLNB)

Compound Figure Separation of Biomedical Images: Mining Large datasets for Self-supervised Learning

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Yao, Tianyuan Qu, Chang Long, Jun Liu, Quan Deng, Ruining Tian, Yuanhan Xu, Jiachen Jha, Aadarsh Asad, Zuhayr Bao, Shunxing Zhao, Mengyang Fogo, Agnes B. Landman, Bennett A. Yang, Haichun Chang, Catie Huo, Yuankai Vanderbilt University Department of Computer Science NashvilleTN37215 United States Central South University Big Data Institute Hunan Changsha410083 China Vanderbilt University Department of Electrical and Computer Engineering NashvilleTN37215 United States Dartmouth College HanoverNH03755 United States Vanderbilt University Medical Center Department of Pathology NashvilleTN37215 United States

With the rapid development of self-supervised learning (e.g., contrastive learning), the importance of having large-scale images (even without annotations) for training a more generalizable AI model has been widely recognized in medical image analysis. However, collecting large-scale task-specific unannotated data at scale can be challenging for individual labs. Existing online resources, such as digital books, publications, and search engines, provide a new resource for obtaining large-scale images. However, published images in healthcare (e.g., radiology and pathology) consist of a considerable amount of compound figures with subplots. In order to extract and separate compound figures into usable individual images for downstream learning, we propose a simple compound figure separation (SimCFS) framework without using the traditionally required detection bounding box annotations, with a new loss function and a hard case simulation. Our technical contribution is four-fold: (1) we introduce a simulation-based training framework that minimizes the need for resource extensive bounding box annotations;(2) we propose a new side loss that is optimized for compound figure separation;(3) we propose an intra-class image augmentation method to simulate hard cases;and (4) to the best of our knowledge, this is the first study that evaluates the efficacy of leveraging self-supervised learning with compound image separation. From the results, the proposed SimCFS achieved state-of-the-art performance on the ImageCLEF 2016 Compound Figure Separation database. The pretrained self-supervised learning model using large-scale mined figures improved the accuracy of downstream image classification tasks with a contrastive learning algorithm. The source code of SimCFS is made publicly available at https://***/hrlblab/ImageSeperation. © 2022, CC BY.

关键词： Supervised learning

DP2-Pub: Differentially Private High-Dimensional data Publication with Invariant Post Randomization

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Jiang, Honglu Yu, Haotian Cheng, Xiuzhen Pei, Jian Pless, Robert Yu, Jiguo The Department of Computer Science and Software Engineering Miami University OxfordOH45056 United States The Department of Computer Science The George Washington University WashingtonDC20052 United States The Department of Data Analytics The George Washington University WashingtonDC20052 United States The School of Computer Science and Technology Shandong University Qingdao266510 China The Departments of Computer Science Biostatistics and Bioinformatics Electrical and Computer Engineering Duke University DurhamNC27708 United States Big Data Institute Qilu University of Technology Jinan250353 China Shandong Fundamental Research Center for Computer Science Qilu University of Technology Shandong Jinan250353 China

A large amount of high-dimensional and heterogeneous data appear in practical applications, which are often published to third parties for data analysis, recommendations, targeted advertising, and reliable predictions. However, publishing these data may disclose personal sensitive information, resulting in an increasing concern on privacy violations. Privacy-preserving data publishing has received considerable attention in recent years. Unfortunately, the differentially private publication of high dimensional data remains a challenging problem. In this paper, we propose a differentially private high-dimensional data publication mechanism (DP2-Pub) that runs in two phases: a Markov-blanket-based attribute clustering phase and an invariant post randomization (PRAM) phase. Specifically, splitting attributes into several low-dimensional clusters with high intra-cluster cohesion and low inter-cluster coupling helps obtain a reasonable allocation of privacy budget, while a double-perturbation mechanism satisfying local differential privacy facilitates an invariant PRAM to ensure no loss of statistical information and thus significantly preserves data utility. We also extend our DP2-Pub mechanism to the scenario with a semi-honest server which satisfies local differential privacy. We conduct extensive experiments on four real-world datasets and the experimental results demonstrate that our mechanism can significantly improve the data utility of the published data while satisfying differential privacy. Copyright © 2022, The Authors. All rights reserved.

关键词： Bayesian networks

LDGCN: An Edge-End Lightweight Dual GCN Based on Single-Channel EEG for Driver Drowsiness Monitoring

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Huang, Jingwei Wang, Chuansheng Huang, Jiayan Fan, Haoyi Grau, Antoni Zhang, Fuquan The College of Computer and Big Data Fuzhou University Fuzhou China The Department of Automatic Control Polytechnic University of Catalonia Barcelona Spain The New Engineering Industry College Putian University Putian China School of Computer and Artificial Intelligence Zhengzhou University China Fujian Provincial Key Laboratory of Information Processing and Intelligent Control Minjiang University Fuzhou China College of Computer and Control Engineering Minjiang University Fuzhou China Key Laboratory of Sichuan Province Sichuan Conservatory of Music Chengdu China Fuzhou Technology Innovation Center of Intelligent Manufacturing information System Minjiang University Fuzhou China Fujian Polytechnic Normal University Fujian Province University Fuzhou China

Driver drowsiness electroencephalography (EEG) signal monitoring can timely alert drivers of their drowsiness status, thereby reducing the probability of traffic accidents. Graph convolutional networks (GCNs) have shown significant advancements in processing the non-stationary, time-varying, and non-Euclidean nature of EEG signals. However, the existing single-channel EEG adjacency graph construction process lacks interpretability, which hinders the ability of GCNs to effectively extract adjacency graph features, thus affecting the performance of drowsiness monitoring. To address this issue, we propose an edge-end lightweight dual graph convolutional network (LDGCN). Specifically, we are the first to incorporate neurophysiological knowledge to design a Baseline Drowsiness Status Adjacency Graph (BDSAG), which characterizes driver drowsiness status. Additionally, to express more features within limited EEG data, we introduce the Augmented Graph-level Module (AGM). This module captures global and local information at the graph level, ensuring that BDSAG features remain intact while enhancing effective feature expression capability. Furthermore, to deploy our method on the fourth-generation Raspberry Pi, we utilize Adaptive Pruning Optimization (APO) on both channels and neurons, reducing inference latency by almost half. Experiments on benchmark datasets demonstrate that LDGCN offers the best trade-off between monitoring performance and hardware resource utilization compared to existing state-of-the-art algorithms. All our source code can be found at https://***/BryantDom/Driver-Drowsiness-Monitoring. Copyright © 2024, The Authors. All rights reserved.

关键词： Electroencephalography

Secure and Privacy-preserving data-sharing Framework based on Blockchain Technology for Al-Najaf/Iraq Oil Refinery 19

学校读者我要写书评

暂无评论

Secure and Privacy-preserving Data-sharing Framework based o...

2022 IEEE SmartWorld, 19th IEEE International Conference on Ubiquitous Intelligence and Computing, 2022 IEEE International Conference on Autonomous and Trusted Vehicles Conference, 22nd IEEE International Conference on Scalable Computing and Communications, 2022 IEEE International Conference on Digital Twin, 8th IEEE International Conference on Privacy Computing and 2022 IEEE International Conference on Metaverse, SmartWorld/UIC/ATC/ScalCom/DigitalTwin/PriComp/Metaverse 2022

作者： Umran, Samir M. Lu, SongFeng Abduljabbar, Zaid Ameen Lu, Zhi Feng, Bingyan Zheng, Lu Huazhong University of Science and Technology Hubei Key Laboratory of Distributed System Security Hubei Engineering Research Center on Big Data Security School of Cyber Science and Engineering Wuhan430074 China Iraqi Cement State Company Ministry of Industry and Minerals Baghdad10011 Iraq Shenzhen Huazhong University of Science and Technology Research Institute Shenzhen518057 China University of Basrah College of Education for Pure Sciences Iraq Al-Kunooze University College Technical Computer Engineering Department Basrah Iraq Huazhong University of Science and Technology School of Cyber Science and Engineering Wuhan430074 China Industrial Internet Research Institute Wuhan Huazhong Numerical Control Co. Ltd Wuhan430074 China South-Central University for Nationalities College of Computer Science Wuhan430074 China

ISBN: (纸本)9798350346558

The Industrial Internet of Things or Industry 4.0 efficiently enhances the manufacturing process in terms of raising productivity, system performance, cost reduction, and building large-scale systems. It enables the connection of numerous heterogeneous devices and sensors into the internet network through the utilization of revolutionary techniques that transfer the manufacturing paradigm to intelligent processes. The traditional industrial internet of things adopts centralized architectures and complex encryption algorithms that suffer from several cyber-security threats and lead to data leakage, privacy disclosure, and high computational and communicational costs. In our work, we proposed a decentralized, efficient, low-power, scalable, secure, privacy-preserving, and trusting architecture based on a private blockchain network/smart contract and interplanetary file system technology for the Al-NajafyIraq oil refinery factory. That enables autonomous working and provides a high level of security authentication and privacy preservation, P2P communication, remote access, immutability and information backtracking. Traditional blockchain technology has storage limitation issues, most recent works adopted external servers or the cloud for storage purposes. Our proposed scheme addresses the limitations of blockchain data storage by adopting an interplanetary file system (IPFS) network;provides an efficient data encryption mechanism at the perceptual layer by adopting a lightweight encryption algorithm, high-performance, and ultra-low-power consumption ARM Cortex-M microcontroller. In addition, our architecture succeeds in merging blockchain technology, IoTs, and an IPFS with the oil refinery as an industrial application area. Moreover, it provides an efficient framework that resists common cyber-security attacks and realizes cyber-security requirements represented by data availability, integrity, and confidentiality. © 2022 IEEE.

关键词： Digital storage

Exploiting the Intrinsic Neighborhood Semantic Structure for Domain Adaptation in EEG-based Emotion Recognition

学校读者我要写书评

暂无评论

IEEE Transactions on Affective Computing 2025年

作者： Yang, Yi Wang, Ze Song, Yu Jia, Ziyu Wang, Boyu Jung, Tzyy-Ping Wan, Feng Macau University of Science and Technology Macao Centre for Mathematical Sciences Respiratory Disease AI Laboratory on Epidemic Intelligence and Medical Big Data Instrument Applications Faculty of Innovation Engineering 999078 China Tianjin University of Technology School of Electrical Engineering and Automation Tianjin Key Laboratory of New Energy Power Conversion Transmission and Intelligent Control Tianjin300384 China Chinese Academy of Sciences Beijing Key Laboratory of Brainnetome and Brain-Computer Interface and Brainnetome Center Institute of Automation Beijing100045 China Western University Department of Computer Science Brain Mind Institute LondonONN6A 3K7 Canada University of California at San Diego Swartz Center for Computational Neuroscience Institute for Neural Computation La Jolla CA92093 United States University of Macau Department of Electrical and Computer Engineering Faculty of Science and Technology China University of Macau Centre for Cognitive and Brain Sciences Centre for Artificial Intelligence and Robotics Institute of Collaborative Innovation 999078 China

Due to the inherent non-stationarity and individual differences present in electroencephalogram (EEG) signals, developing a generalizable model that performs well on new subjects is challenging in EEG-based emotion recognition. Most existing domain adaptation (DA) methods typically mitigate these discrepancies by aligning the marginal distributions of domain feature representations. However, when there is a significant difference in the class-conditional distribution between domain features and labels, the domain-invariant features learned by aligning marginal distributions may have limited discriminative ability for unlabeled target instances or even prove counterproductive. To address this issue, we propose a Neighborhood Semantic Aware Learning-based Dynamic Graph Attention Convolution (NSAL-DGAT) approach that learns target semantic information by considering the inter-domain semantic topological structure, thereby improving classifier adaptation for target instances. Specifically, the proposed NSAL framework is designed to capitalize on the insight that after domain feature alignment, some target samples and their neighboring source samples exhibit similar semantics. By leveraging the neighborhood topological structure, we extract and incorporate semantic target features to train a more transferable classifier. Besides, we implement an entropy weighting mechanism to emphasize representative target semantic information, encouraging target instances to prioritize high-confidence individuals within the source neighborhood. We have conducted extensive experiments on the public SEED dataset and our collected the Hearing-Impaired EEG dataset (HIED). The experimental results underscore the efficacy of our proposed NSAL-DGAT approach, showcasing state-of-the-art accuracy in subject-dependent as well as subject-independent scenarios. The source code is available at https://***/YYingDL/NSAL-DGAT. © 2010-2012 IEEE.

关键词： Speech analysis