检索结果-内蒙古大学图书馆

Tongue Color Classification Based on Convolutional Neural Network

学校读者我要写书评

暂无评论

Tongue Color Classification Based on Convolutional Neural Ne...

Future of Information and Communication Conference, FICC 2021

作者： Guangyu, Niu Caiqun, Wang Bo, Yan Yong, Pan Beijing Friendship Hospital Affiliated to the Capital University of Medical Sciences No.95 Yong’an Road Xicheng District Beijing100050 China Beijing Computing Center Building 3 Beike industry No.7 Fengxian Middle Road Haidian District Beijing100094 China Beijing Institute of Technology No.5 Zhongguancun South Street Haidian District Beijing100081 China

ISBN: (纸本)9783030731021

Tongue color classification plays an important role in traditional Chinese medicine. Tongue color is closely related to the physical condition of patients, so it can help doctors to diagnose patients accurately. However, it is difficult to distinguish between different tongue colors. Therefore, it is necessary to develop an effective method to extract high-dimensional tongue color features. Based on deep learning, this paper proposes a new method to improve the accuracy of tongue color classification. Firstly, the semantic convolutional neural network (CNN) is used to extract the tongue image from the background. Then the CNN model is used to extract the tongue color features, and the center loss function is used to enhance the feature discrimination during the training. Experimental results of different verification indexes show that the accuracy of tongue color classification can be improved by the semantic based CNN and center loss function. © 2021, The Author(s), under exclusive license to Springer Nature Switzerland AG.

关键词： Semantic Segmentation

Ultrasound Lung Aeration Map via Physics-Aware Neural Operators

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Wang, Jiayun Ostras, Oleksii Sode, Masashi Tolooshams, Bahareh Li, Zongyi Azizzadenesheli, Kamyar Pinton, Gianmarco Pinton, Gianmarco F. Anandkumar, Anima Department of Computing and Mathematical Sciences California Institute of Technology 1200 E California Blvd PasadenaCA91125 United States Department of Biomedical Engineering University of North Carolina at Chapel Hill 103 South Building Chapel HillNC27514 United States Department of Biomedical Engineering North Carolina State University Campus Box 7625 RaleighNC27695 United States NVIDIA 2788 San Tomas Express Way Santa ClaraCA95051 United States

Lung ultrasound is a growing modality in clinics for diagnosing and monitoring acute and chronic lung diseases due to its low cost and accessibility. Lung ultrasound works by emitting diagnostic pulses, receiving pressure waves and converting them into radio frequency (RF) data, which are then processed into B-mode images with beamformers for radiologists to interpret. However, unlike conventional ultrasound for soft tissue anatomical imaging, lung ultrasound interpretation is complicated by complex reverberations from the pleural interface caused by the inability of ultrasound to penetrate air. The indirect B-mode images make interpretation highly dependent on reader expertise, requiring years of training, which limits its widespread use despite its potential for high accuracy in skilled hands. To address these challenges and democratize ultrasound lung imaging as a reliable diagnostic tool, we propose Luna (the Lung Ultrasound Neural operator for Aeration), an AI model that directly reconstructs lung aeration maps from RF data, bypassing the need for traditional beamformers and indirect interpretation of B-mode images. Luna uses a Fourier neural operator, which processes RF data efficiently in Fourier space, enabling accurate reconstruction of lung aeration maps. From reconstructed aeration maps, we calculate lung percent aeration, a key clinical metric, offering a quantitative, reader-independent alternative to traditional semi-quantitative lung ultrasound scoring methods. The development of Luna involves synthetic and real data: We simulate synthetic data with an experimentally validated approach and scan ex vivo swine lungs as real data. Trained on abundant simulated data and fine-tuned with a small amount of real-world data, Luna achieves robust performance, demonstrated by an aeration estimation error of 9% in ex-vivo swine lung scans. We demonstrate the potential of directly reconstructing lung aeration maps from RF data, providing a foundation for improving

关键词： Beamforming

ANNEALED POTTS MODELS ON RANK-1 INHOMOGENEOUS RANDOM GRAPHS

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Giardinà, Cristian Giberti, Claudio van der Hofstad, Remco Janssen, Guido Maitra, Neeladri University of Modena and Reggio Emilia via G. Campi 213/b Modena41125 Italy University of Modena and Reggio Emilia via G. Amendola 2 Pad. Morselli Reggio E. 42122 Italy Eindhoven University of Technology P.O. Box 513 Eindhoven5600 MB Netherlands Department of Mathematics University of Illinois at Urbana-Champaign 69.2.1 Computing Applications Building 605 East Springfield Avenue ChampaignIL61820 United States

In this paper, we study the annealed ferromagnetic q-state Potts model on sparse rank-1 random graphs, where vertices are equipped with a vertex weight, and the probability of an edge is proportional to the product of the vertex weights. In an annealed system, we take the average on both numerator and denominator of the ratio defining the Boltzmann-Gibbs measure of the Potts model. We show that the thermodynamic limit of the pressure per particle exists for rather general vertex weights. In the infinite-variance weight case, we show that the critical temperature equals infinity. For finite-variance weights, we show that, under a rather general condition, the phase transition is first order for all q ≥ 3. However, we cannot generally show that the discontinuity of the order parameter is unique. We prove this uniqueness under a reasonable condition that holds for various distributions, including uniform, gamma, log-normal, Rayleigh and Pareto distributions. Further, we show that the first-order phase transition persists even for some small positive external field. In the rather relevant case of Pareto distributions with power-law exponent τ, remarkably, the phase transition is first order when τ ≥ 4, but not necessarily when the weights have an infinite third-moment, i.e., when τ ∈ (3, 4). More precisely, the phase transition is second order for τ ∈ (3, τ(q)], while it is first order when τ > τ(q), where we give an explicit equation that τ(q) *** Codes 82B20 (Primary) 82B26, 05C80 (Secondary) © 2025, CC BY-NC-ND.

关键词： Potts model

Diagnosis of COVID-19 Pneumonia via a Novel Deep Learning Architecture

学校读者我要写书评

暂无评论

Journal of Computer Science & technology 2022年第2期37卷 330-343页

作者： Xin Zhang Siyuan Lu Shui-Hua Wang Xiang Yu Su-Jing Wang Lun Yao Yi Pan Yu-Dong Zhang Department of Medical Imaging The Fourth People's Hospital of Huai'anHuai'an 223002China School of Informatics University of LeicesterLeicesterLE17RHU.K. School of Architecture Building and Civil Engineering Loughborough UniversityLoughboroughLE113TUU.K. School of Mathematics and Actuarial Science University of LeicesterLeicesterLE17RHU.K. Key Laboratory of Behavior Sciences Institute of PsychologyChinese Academy of SciencesBeijing 100101China Department of Psychology University of the Chinese Academy of SciencesBeijing 100101China Department of Infection Diseases The Fourth People's Hospital of Huai'anHuaian 223002China Department of Computer Science Georgia State UniversityAtlanta 30302-5060U.S.A. Department of Information Systems Faculty of Computing and Information TechnologyKing Abdulaziz University Jeddah 21589Saudi Arabia

COVID-19 is a contagious infection that has severe effects on the global economy and our daily *** diagnosis of COVID-19 is of importance for consultants,patients,and *** this study,we use the deep learning network AlexNet as the backbone,and enhance it with the following two aspects:1)adding batch normalization to help accelerate the training,reducing the internal covariance shift;2)replacing the fully connected layer in AlexNet with three classifiers:SNN,ELM,and ***,we have three novel models from the deep COVID network(DC-Net)framework,which are named DC-Net-S,DC-Net-E,and DC-Net-R,*** comparison,we find the proposed DC-Net-R achieves an average accuracy of 90.91%on a private dataset(available upon email request)comprising of 296 images while the specificity reaches 96.13%,and has the best performance among all three proposed *** addition,we show that our DC-Net-R also performs much better than other existing algorithms in the literature.

关键词： pneumonia COVID-19 convolutional neural network AlexNet deep learning

SAIBench: Benchmarking AI for Science

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Li, Yatao Zhan, Jianfeng Institute of Computing Technology Chinese Academy of Science No.6 Kexueyuan South RoadHaidian District Beijing100190 China Yuquan Road Shijingshan District Beijing100049 China Microsoft Research Asia Building 2 No. 5 Dan Ling Street Haidian District Beijing100080 China

Scientific research communities are embracing AI-based solutions to target tractable scientific tasks and improve research work-flows. However, the development and evaluation of such solutions are scattered across multiple disciplines. We formalize the problem of scientific AI benchmarking, and propose a system called SAIBench in the hope of unifying the efforts and enabling low-friction on-boarding of new disciplines. The system approaches this goal with SAIL, a domain-specific language to decouple research problems, AI models, ranking criteria, and software/hardware configuration into reusable modules. We show that this approach is flexible and can adapt to problems, AI models, and evaluation methods defined in different perspectives. The project homepage is https://***/SAIBench. © 2022, CC BY-NC-ND.

关键词： Benchmarking

Survey on Large Language Model-Enhanced Reinforcement Learning: Concept, Taxonomy, and Methods

学校读者我要写书评

暂无评论

IEEE Transactions on Neural Networks and Learning Systems 2024年第6期36卷 9737-9757页

作者： Yuji Cao Huan Zhao Yuheng Cheng Ting Shu Yue Chen Guolong Liu Gaoqi Liang Junhua Zhao Jinyue Yan Yun Li Department of Mechanical and Automation Engineering The Chinese University of Hong Kong Hong Kong SAR China Department of Building Environment and Energy Engineering The Hong Kong Polytechnic University Hong Kong China School of Science and Engineering The Chinese University of Hong Kong Shenzhen China Center for Crowd Intelligence Shenzhen Institute of Artificial Intelligence and Robotics for Society (AIRS) Shenzhen China National Engineering Laboratory for Big Data System Computing Technology Shenzhen University Shenzhen China School of Electrical and Electronic Engineering Nanyang Technological University Jurong West Singapore School of Mechanical Engineering and Automation Harbin Institute of Technology Shenzhen China Shenzhen Institute for Advanced Study University of Electronic Science and Technology of China Shenzhen China i4AI Ltd. London U.K.

With extensive pretrained knowledge and high-level general capabilities, large language models (LLMs) emerge as a promising avenue to augment reinforcement learning (RL) in aspects, such as multitask learning, sample efficiency, and high-level task planning. In this survey, we provide a comprehensive review of the existing literature in LLM-enhanced RL and summarize its characteristics compared with conventional RL methods, aiming to clarify the research scope and directions for future studies. Utilizing the classical agent-environment interaction paradigm, we propose a structured taxonomy to systematically categorize LLMs’ functionalities in RL, including four roles: information processor, reward designer, decision-maker, and generator. For each role, we summarize the methodologies, analyze the specific RL challenges that are mitigated and provide insights into future directions. Finally, the comparative analysis of each role, potential applications, prospective opportunities, and challenges of the LLM-enhanced RL are discussed. By proposing this taxonomy, we aim to provide a framework for researchers to effectively leverage LLMs in the RL field, potentially accelerating RL applications in complex applications, such as robotics, autonomous driving, and energy systems.

关键词： Taxonomy Surveys Visualization Reviews Planning Reinforcement learning Games Transformers Natural language processing Large language models

A thermodynamic approach to optimization in complex quantum systems

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Imparato, Alberto Chancellor, Nicholas De Chiara, Gabriele Department of Physics and Astronomy Aarhus University Denmark Department of Physics Joint Quantum Centre Durham-Newcastle Durham University Newcastle University School of Computing Urban Science Building NewcastleNE4 5TG United Kingdom Centre for Quantum Materials and Technology School of Mathematics and Physics Queen's University Belfast BelfastBT7 1NN United Kingdom

We consider the problem of finding the energy minimum of a complex quantum Hamiltonian by employing a non-Markovian bath prepared in a low energy state. The energy minimization problem is thus turned into a thermodynamic cooling protocol in which we repeatedly put the system of interest in contact with a colder auxiliary system. By tuning the internal parameters of the bath, we show that the optimal cooling is obtained in a regime where the bath exhibits a quantum phase transition in the thermodynamic limit. This result highlights the importance of collective effects in thermodynamic devices. We furthermore introduce a two-step protocol that combines the interaction with the bath with a measure of its energy. While this protocol does not destroy coherence in the system of interest, we show that it can further enhance the cooling effect. Copyright © 2023, The Authors. All rights reserved.

关键词： Phase transitions

Wind speed prediction and insight for generalized predictive modeling framework: a comparative study for different artificial intelligence models

学校读者我要写书评

暂无评论

Neural computing and Applications 2024年第23期36卷 14119-14150页

作者： Bhagat, Suraj Kumar Tiyasha, Tiyasha Shather, A.H. Jamei, Mehdi Kumar, Adarsh Al-Khafaji, Zainab Goliatt, Leonardo Shafik, Shafik S. Alawi, Omer A. Yaseen, Zaher Mundher Faculty of Civil Engineering Ton Duc Thang University Ho Chi Minh City Viet Nam Department of Computer Engineering Technology Al Kitab University Altun Kopru Kirkuk00964 Iraq Faculty of Civil Engineering and Architecture Shahid Chamran University of Ahvaz Ahvaz Iran Canadian Centre for Climate Change and Adaptation University of Prince Edward Island St PetersPE Canada Building and Construction Techniques Engineering Department College of Engineering and Engineering Techniques Al-Mustaqbal University Hillah Babylon51001 Iraq Department of Computational and Applied Mechanics Federal University of Juiz de Fora Juiz de Fora Brazil Experimental Nuclear Radiation Research Group Scientific Research Center Al-Ayen University Thi-Qar64001 Iraq Johor Bahru Skudai81310 Malaysia Civil and Environmental Engineering Department King Fahd University of Petroleum and Minerals Dhahran31261 Saudi Arabia Department of Civil Engineering School of Engineering and Computing Dev Bhoomi Uttarakhand University Dehradun India Dept of Environmental Science Gitam School of Science GITAM University Visakhapatnam India

Wind speed (WS) has played a vital role in local urban and sub-urban weather, agriculture, and ecosystem. Several meteorological parameters are influencing WS such as relative humidity (at 2 m, %), surface pressure (kPa), maximum temperature (at 2 m, °C), minimum temperature (at 2 m, °C), average temperature (at 2 m, °C), and all sky insolation incident on a horizontal surface (kW-h/m2/day). The current research was conducted to predict WS at different locations at Vietnam using the feasibility of computer aid models (i.e., multivariate adaptive regression splines (MARS), extreme gradient boosting (XGBoost) and random forest generator (Ranger)). Pearson correlation (PC) was investigated to select the high significant predictors to predict the WS at 10 m high. All inputs (maximum number, 6) are chosen by the PC approach for PhuongNinh, DaNang, and HaNoi;and for minimum number of inputs i.e four, are selected for PhuongHung, CanTho, and SaPa city;that exhibit the relationship with WS, citywise. The sequence selection of input parameters differed in each station as per the PC analysis. Based on the statistical evaluation and graphical presentation, MARS model attained the best prediction results, followed by XGBoost and Ranger. MARS predictive model remains at the top performance among others based on 95% confidence interval. © The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature 2024.

关键词： Forecasting

A directional regularization method for the limited-angle Helsinki Tomography Challenge using the Core Imaging Library (CIL)

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Jørgensen, Jakob Sauer Papoutsellis, Evangelos Murgatroyd, Laura Fardell, Gemma Pasca, Edoardo Department of Applied Mathematics and Computer Science Technical University of Denmark. Richard Petersens Plads Building 324 Kgs. Lyngby2800 Denmark Department of Mathematics The University of Manchester Oxford Road Alan Turing Building ManchesterM13 9PL United Kingdom Finden Ltd Rutherford Appleton Laboratory Harwell Campus DidcotOX11 0QX United Kingdom Scientific Computing Department Science & Technology Facilities Council Rutherford Appleton Laboratory Harwell Campus DidcotOX11 0QX United Kingdom

This article presents the algorithms developed by the Core Imaging Library (CIL) developer team for the Helsinki Tomography Challenge 2022. The challenge focused on reconstructing 2D phantom shapes from limited-angle computed tomography (CT) data. The CIL team designed and implemented five reconstruction methods using CIL (https://***/cil/), an open-source Python package for tomographic imaging. The CIL team adopted a model-based reconstruction strategy, unique to this challenge with all other teams relying on deep-learning techniques. The CIL algorithms showcased exceptional performance, with one algorithm securing the third place in the competition. The best-performing algorithm employed careful CT data pre-processing and an optimization problem with single-sided directional total variation regularization combined with isotropic total variation and tailored lower and upper bounds. The reconstructions and segmentations achieved high quality for data with angular ranges down to 50 degrees, and in some cases acceptable performance even at 40 and 30 degrees. This study highlights the effectiveness of model-based approaches in limited-angle tomography and emphasizes the importance of proper algorithmic design leveraging on available prior knowledge to overcome data limitations. Finally, this study highlights the flexibility of CIL for prototyping and comparison of different optimization methods.1MSC Codes 65R32, 94A08, 65K10 © 2023, CC BY.

关键词： Computerized tomography