检索结果-内蒙古大学图书馆

Diffusion-empowered AutoPrompt MedSAM

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Huang, Peng Hu, Shu Peng, Bo Gong, Xun Yin, Penghang Zhu, Hongtu Wu, Xi Wang, Xin School of Computing and Artificial Intelligence Southwest Jiaotong University Chengdu China Department of Computer and Information Technology Purdue University United States Department of Mathematics and Statistics University at Albany State University of New York United States University of North Carolina Chapel Hill United States School of Computer Science Chengdu University of Information Technology Chengdu China College of Integrated Health Sciences the AI Plus Institute University at Albany State University of New York United States

MedSAM, a medical foundation model derived from SAM, has achieved remarkable success across various medical domains. Nonetheless, its clinical adoption encounters two primary challenges: the reliance on labor-intensive manual prompt generation, which places a considerable burden on clinicians, and the lack of semantic labeling in the generated segmentation masks for organs or lesions, which hinders its usability for non-expert users. To overcome these limitations, we propose an end-to-end interactive segmentation model, AutoMedSAM. AutoMedSAM employs a diffusion-based class prompt encoder to generate prompt embeddings guided by the prompt class indices and eliminates the reliance on manual prompts. During the diffusion process, the encoder progressively captures the semantic structure and fine-grained features of the target object, injecting semantic information into the prediction pipeline. Furthermore, we introduce an uncertainty-aware joint optimization strategy that integrates pixel-based, region-based, and distribution-based losses. This approach harnesses MedSAM’s pre-trained knowledge and various loss functions to enhance the model’s generalization. Experimental evaluations on diverse datasets demonstrate that AutoMedSAM not only achieves superior segmentation performance but also extends its applicability to clinical environments and nonspecialist users. Code is available at https://github. com/HP-ML/***. Copyright © 2025, The Authors. All rights reserved.

关键词： Semantic Segmentation

DEGSTalk: Decomposed Per-Embedding Gaussian Fields for Hair-Preserving Talking Face Synthesis

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Deng, Kaijun Zheng, Dezhi Xie, Jindong Wang, Jinbao Xie, Weicheng Shen, Linlin Song, Siyang Computer Vision Institute School of Computer Science and Software Engineering Shenzhen University China National Engineering Laboratory for Big Data System Computing Technology Shenzhen University China Guangdong Provincial Key Laboratory of Intelligent Information Processing China Department of Computer Science University of Exeter United Kingdom

Accurately synthesizing talking face videos and capturing fine facial features for individuals with long hair presents a significant challenge. To tackle these challenges in existing methods, we propose a decomposed per-embedding Gaussian fields (DEGSTalk), a 3D Gaussian Splatting (3DGS)-based talking face synthesis method for generating realistic talking faces with long hairs. Our DEGSTalk employs Deformable Pre-Embedding Gaussian Fields, which dynamically adjust pre-embedding Gaussian primitives using implicit expression coefficients. This enables precise capture of dynamic facial regions and subtle expressions. Additionally, we propose a Dynamic Hair-Preserving Portrait Rendering technique to enhance the realism of long hair motions in the synthesized videos. Results show that DEGSTalk achieves improved realism and synthesis quality compared to existing approaches, particularly in handling complex facial dynamics and hair preservation. Our code will be publicly available at https://***/CVI-SZU/DEGSTalk. Copyright © 2024, The Authors. All rights reserved.

关键词： Gaussian distribution

Auction Theory and Game Theory Based Pricing of Edge computing Resources: A Survey

学校读者我要写书评

暂无评论

IEEE Internet of Things Journal 2025年

作者： Yu, Jiguo Liu, Shun Zou, Yifei Wang, Guijuan Hu, Chunqiang University of Electronic Science and Technology of China School of Information and Software Engineering Chengdu610054 China Qilu University of Technology Big Data Institute Jinan250353 China Qufu Normal University School of Computer Science Shandong Rizhao China Shandong University School of Computer Science and Technology Shandong Qingdao266237 China Key Laboratory of Computing Power Network and Information Security Ministry of Education Shandong Computer Science Center Jinan250353 China Shandong Fundamental Research Center for Computer Science Shandong Provincial Key Laboratory of Computer Networks Jinan250353 China Chongqing University School of Big Data and Software Engineering Chongqing400044 China

The proliferation of Internet of Things (IoT) devices and edge computing applications has heightened the demand for efficient resource allocation and pricing mechanisms. Effective pricing strategies play a crucial role in optimizing resource utilization, reducing latency, and ensuring fair resource distribution among users. Both auction theory and game theory have emerged as fundamental economic frameworks for designing pricing mechanisms in edge computing. This paper presents a comprehensive survey of pricing strategies that leverage both auction and game theory, systematically categorizing existing research based on auction mechanisms and game-theoretic models. Unlike previous surveys that focus solely on auction-based pricing, this work integrates game-theoretic perspectives, providing a holistic analysis of strategic interactions in edge computing resource allocation. Furthermore, we discuss key challenges, emerging trends, and future research directions aimed at enhancing efficiency, fairness, and privacy protection in cloud-edge collaborative computing environments. © 2014 IEEE.

关键词： Mobile edge computing

Reducing Network Distance Measurement Overhead: A Tensor Completion Solution With a New Minimum Sampling Bound

学校读者我要写书评

暂无评论

IEEE/ACM Transactions on Networking 2024年

作者： Tian, Jiazheng Xie, Kun Wang, Xin Wen, Jigang Xie, Gaogang Cao, Jiannong Liang, Wei Li, Kenli Hunan University College of Computer Science and Electronics Engineering Changsha410082 China Ministry of Education Key Laboratory of 'Fusion Computing of Supercomputing and Artificial Intelligence' Changsha410082 China State of New York University at Stony Brook Department of Electrical and Computer Engineering Stony BrookNY11794 United States Hunan University of Science and Technology School of Computer Science and Engineering Xiangtan411199 China Chinese Academy of Sciences Computer Network Information Center Beijing100080 China University of Chinese Academy of Sciences School of Computer Science and Technology Beijing101408 China The Hong Kong Polytechnic University Department of Computing Hong Kong Hong Kong

Network distance measurement is crucial for evaluating network performance, attracting significant research attention. However, conducting measurements for the entire network is exceedingly expensive and time-consuming, making the reduction of network distance measurement costs a top priority. The tensor completion method efficiently reduces measurement costs by utilizing a small amount of measured data to estimate the entire network's distance data. Unfortunately, current tensor completion methods still suffer from issues such as complex sample selection, high measurement overhead, slow recovery, and low inference accuracy. To address the aforementioned challenges, we present an online network-wide distance measurement scheme. In this approach, continuous distance data are structured into sliding-window-based tensors. Our method incorporates a lightweight sample selection algorithm with a lowest sampling bound and a rapid, accurate unmeasured data inference algorithm. We have conducted extensive experiments using four real network distance datasets and two citywide crowd flow datasets. The empirical evaluations demonstrate the effectiveness of our approach, particularly in reducing measurement costs and enhancing data recovery accuracy. © 1993-2012 IEEE.

关键词： Distance measurement

Input-driven chaotic dynamics in vortex spin-torque oscillator

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Imai, Yusuke Nakajima, Kohei Tsunegi, Sumito Taniguchi, Tomohiro Research Center for Emerging Computing Technologies Ibaraki Tsukuba305-8568 Japan Graduate School of Information Science and Technology The University of Tokyo Bunkyo-ku Tokyo113-8656 Japan Saitama332-0012 Japan

A new research topic in spintronics relating to the operation principles of brain-inspired computing is input-driven magnetization dynamics in nanomagnet. In this paper, the magnetization dynamics in a vortex spin-torque oscillator (STO) driven by a series of random magnetic field are studied through a numerical simulation of the Thiele equation. It is found that input-driven synchronization occurs in the weak perturbation limit, as found recently. As well, chaotic behavior is newly found to occur in the vortex core dynamics for a wide range of parameters, where synchronized behavior is disrupted by an intermittency. Ordered and chaotic dynamical phases are examined by evaluating the Lyapunov exponent. The relation between the dynamical phase and the computational capability of physical reservoir computing is also studied. © 2023, CC BY.

关键词： Vortex flow

A Classification of Inflammatory Bowel Disease using Ensemble Learning Model

学校读者我要写书评

暂无评论

A Classification of Inflammatory Bowel Disease using Ensembl...

Software, Systems and information technology (SSITCON), International Conference on

作者： Surendar Rama Sitaraman Myasar Mundher Adnan K. Maharajan R Krishna Prakash R. Dhilipkumar Intel Corporation California USA Department of Computers Techniques Engineering College of Technical Engineering The Islamic University Najaf Iraq Department of Computer Science and Engineering School of Computing Kalasalingam Academy of Research and Education Krishnankoil India Department of Information Technology Sona College of Technology Salem India Department of Electronics and Communication Engineering New Prince Shri Bhavani College of Engineering and Technology Chennai India

ISBN: (数字)9798350352931

ISBN: (纸本)9798350352948

Inflammatory Bowel Disease (IBD) is an enduring, intermittent bowel disorder with unidentified mechanism and etiology. It typically contains Ulcerative colitis (UC) and Crohn’s disease (CD) that are categorized by several bowel ulcers. An impedance phase displaces dynamic development, accumulating susceptibility to exterior pathogens and theoretically leading to neurotic symptoms which encompasses bowel blockade, intestinal tear and liver lesions, with high disability ratio. The existing models have more overfitting, and class imbalance problems. To overcome that, an Ensemble Learning (EL) model such as Logistic Regression, Random Forest and Gaussian Naive Bayes is proposed in this research. Initially, the collected Gene Expression Series 106817 (GSE106817) dataset is preprocessed with Recurrent MultiArray Average (RMA) Normalization and Log2 Data Transformation to eliminate the noises. Next, Recursive Feature Elimination (RFE) technique is used to reduce high-dimensional data for selecting optimal features; and finally, the proposed EL model gives a precise rate in detecting and classifying IBD as CD, UC and Healthy controls. The proposed EL model gives better results than the other compared existing methods Random Forest-Support Vector Machine (RF-SVM) in the context of recall, accuracy, f1 score, and precision as 99.38%, 99.75%, 94.72%, and 97.94% respectively.

关键词： Accuracy Sensitivity Noise Feature extraction Software systems Vectors Ensemble learning Random forests Overfitting Ulcerative colitis

Estimation of Clinical Parameters for Glaucoma Detection Using Image Processing and Deep Learning Techniques from the Biomedical Human Eye Images

学校读者我要写书评

暂无评论

Estimation of Clinical Parameters for Glaucoma Detection Usi...

Emerging Innovations and Advanced computing (INNOCOMP), International Conference on

作者： Guru Prasad M S Chandrappa S Gulshan Dhasmana Naveen Kumar H N Shilpa Mathpati Ninad Gangodkar Dept. of Computer Science and Engg. Graphic Era (Deemed to be University) Dehradun India Dept. of Computer Science and Engneering Jain (Deemed-to-be University) Bengaluru India Department of ECE Vidyavardhaka College of Engineering Mysuru India School of Computing and Information Technology REVA University Bengaluru India

ISBN: (数字)9798350376470

ISBN: (纸本)9798350376487

The technique for distinguishing the essential glaucoma is being exhibited in this segment utilizing a changed neural system with the back spread calculation ideas. The morphological administrators ideas are being utilized for the handling of the cup and disc, in the sense, the cup and the disc zones are discovered from which the proportion is registered, from where the illness can be recognition seeing the proportion. Utilizing an updated highlight extraction process, the highlights of the disc portion and the cup can be discovered, which could be utilized in the idea of cup to plate apportion (cup to disc ratio) recognition, regardless of whether the given test image is glaucomatic or not. Classification is also done using artificial neural networks (ANN) along with the training process. Simulations have been done in the Matlab environment by developing. m codes. Once the developed program is run, the simulation results are observed. From the simulation results, it can be observed that using the computed values of the cup to disc ratio, the severity of the disease can be computed (whether it is normal or moderate or severe glaucoma). Databases have been taken from the hospitals and on-line standard websites. The quantitative results are tabulated in a table. Finally, it was found that after taking a survey of 50 images, 16 were affected with normal, 16 were affected with moderate and the rest were affected with severe case of glaucoma. The Matlab results of the simulation done using coding demonstrated the profoundness with respect to the strategy proposed or developed in this exploration write-up using the concept of neural networks.

关键词： Glaucoma Surveys Training Technological innovation Simulation Artificial neural networks Standards

TransXNet: Learning Both Global and Local Dynamics with a Dual Dynamic Token Mixer for Visual Recognition

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Lou, Meng Zhang, Shu Zhou, Hong-Yu Yang, Sibei Wu, Chuan Yu, Yizhou School of Computing and Data Science The University of Hong Kong Hong Kong AI Lab Deepwise Healthcare Beijing China Department of Biomedical Informatics Harvard Medical School Boston United States Department of Computer Science The University of Hong Kong Hong Kong School of Information Science and Technology ShanghaiTech University Shanghai China

Recent studies have integrated convolutions into transformers to introduce inductive bias and improve generalization performance. However, the static nature of conventional convolution prevents it from dynamically adapting to input variations, resulting in a representation discrepancy between convolution and self-attention as the latter computes attention maps dynamically. Furthermore, when stacking token mixers that consist of convolution and self-attention to form a deep network, the static nature of convolution hinders the fusion of features previously generated by self-attention into convolution kernels. These two limitations result in a sub-optimal representation capacity of the entire network. To find a solution, we propose a lightweight Dual Dynamic Token Mixer (D-Mixer) to simultaneously learn global and local dynamics via computing input-dependent global and local aggregation weights. D-Mixer works by applying an efficient global attention module and an input-dependent depthwise convolution separately on evenly split feature segments, endowing the network with strong inductive bias and an enlarged receptive field. We use D-Mixer as the basic building block to design TransXNet, a novel hybrid CNN-Transformer vision backbone network that delivers compelling performance. In the ImageNet-1K classification, TransXNet-T surpasses Swin-T by 0.3% in top-1 accuracy while requiring less than half of the computational cost. Furthermore, TransXNet-S and TransXNet-B exhibit excellent model scalability, achieving top-1 accuracy of 83.8% and 84.6% respectively, with reasonable computational costs. Additionally, our proposed network architecture demonstrates strong generalization capabilities in various dense prediction tasks, outperforming other state-of-the-art networks while having lower computational costs. Code is publicly available at https://***/LMMMEng/TransXNet. © 2023, CC BY.

关键词： Mixers (machinery)

Predictive Classification of Covid-19: Assessing the Impact of Digital Technologies

学校读者我要写书评

暂无评论

Predictive Classification of Covid-19: Assessing the Impact ...

International conference of Electronics, Communication and Aerospace technology (ICECA)

作者： Bonda Kiran Kumar Anurima Majumdar Saidatul Akmar Ismail Rohit R Dixit Hairani Wahab Md Hafizi Ahsan KL School of Architecture KL Deemed to be University Koneru Lakshmaiah Educational Foundation vaddeswaram Andhra Pradesh Department of Electronics and Communication Engineering Guru Nanak Institute of Technology School of Information Science College of Computing Informatics & Mathematics Universiti Teknologi MARA Shah Alam Malaysia Siemens Healthineers School of Information Science College of Computing Informatics and Mathematics Universiti Teknologi MARA Labuan Faculty of International Finance Universiti Malaysia Sabah

The recent severe outbreak of the disease has been linked to the COVID-19 virus. It is essential for healthcare professionals and individuals with the capacity to provide intensive care to offer clear explanations of this situation. To address this issue, the World Health Organization (WHO) has implemented several measures. The National Emergency Telecritical Care Network (NETCCN) has the capability to assess these actions using digital technologies. This section provides a comprehensive explanation of these developments, and NETCCN has put forth a recommendation. Version 2.0 of this digital health data will be made available; it will significantly enhance potential crises in the future. Artificial Intelligence (AI) is used in the suggested digital prediction approach to gather past data on the infection. Telemetry, sophisticated maintenance robotics, and monitoring intelligence may all be found with this technology. These may be created to incorporate data relevant to various infections. Health experts will coordinate all of these. To maintain sustained preparedness for emergency catastrophe protection, frequent testing should be conducted.

关键词：