检索结果-内蒙古大学图书馆

IEEE Transactions on Pattern Analysis and Machine Intelligence 2025年第7期47卷 5538-5555页

作者： Yan, Kun Wei, Fangyun Dai, Shuyu Wu, Minghui Wang, Ping Xu, Chang Peking University School of Computer Science Beijing100871 China University of Sydney School of Computer Science DarlingtonNSW2008 Australia Peking University School of Software and Microelectronics Beijing100871 China Shanghai Artificial Intelligence Laboratory and Mininglamp Technology China Peking University National Engineering Research Center for Software Engineering Beijing100871 China Peking University School of Software and Microelectronics China China

Prior research in video object segmentation (VOS) predominantly relies on videos with dense annotations. However, obtaining pixel-level annotations is both costly and time-intensive. In this work, we highlight the potential of effectively training a VOS model using remarkably sparse video annotations - specifically, as few as one or two labeled frames per training video, yet maintaining near equivalent performance levels. We introduce this innovative training methodology as low-shot video object segmentation, abbreviated as low-shot VOS. Central to this method is the generation of reliable pseudo labels for unlabeled frames during the training phase, which are then used in tandem with labeled frames to optimize the model. Notably, our strategy is extremely simple and can be incorporated into the vast majority of current VOS models. For the first time, we propose a universal method for training VOS models on one-shot and two-shot VOS datasets. In the two-shot configuration, utilizing just 7.3% and 2.9% of labeled data from the YouTube-VOS and DAVIS benchmarks respectively, our model delivers results on par with those trained on completely labeled datasets. It is also worth noting that in the one-shot setting, a minor performance decrement is observed in comparison to models trained on fully annotated datasets. Code and models are available at https://***/yk-pku/Low-shot-VOS. © 1979-2012 IEEE.

关键词： Video analysis

来源：评论

学校读者我要写书评

暂无评论

Enhancing Large Language Model Inference Efficiency via Lookahead Cache Filtering

Enhancing Large Language Model Inference Efficiency via Look...

引用

2025 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2025

作者： Ou, Jie Chen, Yueming Jiang, Shuaihong Tian, Wenhong School of Information and Software Engineering University of Electronic Science and Technology of China Chengdu China

ISBN: (纸本)9798350368741

The large Key-Value (KV) cache is a significant challenge in deploying Large Language Models (LLMs). Current research addressing these issues employs cache compression techniques, which we find suffer from information loss and the "lost-in-the-middle" problem. We propose the Lookahead Cache Filtering (LCF), which retains the full cache in host memory to keep complete details, using the Important Key-Value Lookahead Prediction by Approximate Sorting and the Gather-based Matrix Multiplication on CPU, to filter important KV cache to reduce the overhead of loading cache to GPU and improve the throughput of inference. Finally, we use the Multi-Scale Pyramid Information Fusion mechanism to enhance information fusion to further improve the effectiveness of inference results. Experiments demonstrate that LCF effectively maintains accuracy without introducing additional latency while reducing memory requirements. Our code will be available on GitHub. © 2025 IEEE.

关键词： Approximate Sorting Important Key-Value Lookahead Information Fusion Lost-in-the-Middle Operator Fusion on CPU

来源：评论

学校读者我要写书评

暂无评论

A Comprehensive Survey of Techniques for Lung Cancer Diagnosis and Prediction 7

A Comprehensive Survey of Techniques for Lung Cancer Diagnos...

引用

7th Edition Global Conference on Wireless and Optical Technologies, GCWOT 2024

作者： Jamil, Danish Shah, Syed Mehr Ali Al-Jarwani, Fatma Mahfoodh Mohammed Sir Syed University of Engineering & Technology Department of Software Engineering Karachi Pakistan College of Computing and Information Sciences University of Technology and Applied Sciences Musandam Oman

ISBN: (纸本)9798331534271

Despite significant advancements, lung cancer remains a formidable global health challenge, necessitating effective diagnostic and prognostic methodologies. This survey paper examines current literature to identify and evaluate methods and models utilized for lung cancer diagnosis and prognosis, aiming to enhance early detection strategies for improved patient outcomes. The research addresses critical questions: What are the current methods and models employed for lung cancer diagnosis and prognosis, their strengths, limitations, and potential for future improvement? It investigates specific applications of deep learning architectures such as CNN, GoogleNet, VGG-16, U-Net, and machine learning algorithms like XGBoost, SVM, KNN, ANN, Random Forest, and hybrid models in lung cancer detection. Methodologically, a systematic review across key databases using relevant keywords was conducted to synthesize findings. The study highlights the strengths and limitations of existing approaches and identifies gaps in model interpretability, real-world validation, and integration of diverse techniques. Conclusions underscore the transformative potential of advanced computational methods in enhancing lung cancer diagnosis and prognosis, proposing avenues for future research to address current challenges and improve patient care. © 2024 IEEE.

关键词： Lung cancer

来源：评论

学校读者我要写书评

暂无评论

A Formal Design of Parallel CRC Circuit 5

A Formal Design of Parallel CRC Circuit

引用

5th International Conference on Electronic engineering and Informatics, EEI 2023

作者： Nie, Yuchen Cai, Fangfei Zhang, Keran Zhong, Sheng Luo, Hangzai Xi'an Microelectronics Technology Institute Department of Software Engineering Xi'an710065 China Northwest University Department of Software Engineering Xi'an710127 China

ISBN: (纸本)9798350327076

Cyclic Redundancy Codes (CRCs) are useful to detect burst errors in storage and communications systems. This paper analyzes the theoretical principle of CRC and the parallel CRC structure. Besides, this paper proposes a design method for the parallel CRC, which can generate parallel CRC circuits with different parallel bit width, different CRC generating polynomial, and different input data length. The FPGA performance (LUT consumption and max frequency) of our approach is comparable to the state-of-art parallel CRC method. This paper may be useful for parallel CRC circuit design. © 2023 IEEE.

关键词： Field programmable gate arrays (FPGA)

来源：评论

学校读者我要写书评

暂无评论

Fake News Classification: Past, Current, and Future

引用

Computers, Materials & Continua 2023年第11期77卷 2225-2249页

作者： Muhammad Usman Ghani Khan Abid Mehmood Mourad Elhadef Shehzad Ashraf Chaudhry Department of Computer Science University of Engineering and TechnologyLahore54890Pakistan Department of Computer Science&Information Technology Abu Dhabi UniversityAbu Dhabi59911United Arab Emirates Department of Software Engineering Faculty of Engineering and ArchitectureNisantasi UniversityIstanbulTurkey

The proliferation of deluding data such as fake news and phony audits on news web journals,online publications,and internet business apps has been aided by the availability of the web,cell phones,and social *** can quickly fabricate comments and news on social *** most difficult challenge is determining which news is real or ***,tracking down programmed techniques to recognize fake news online is *** an emphasis on false news,this study presents the evolution of artificial intelligence techniques for detecting spurious social media *** study shows past,current,and possible methods that can be used in the future for fake news *** different publicly available datasets containing political news are utilized for performing *** supervised learning algorithms are used,and their results show that conventional Machine Learning(ML)algorithms that were used in the past perform better on shorter text *** contrast,the currently used Recurrent Neural Network(RNN)and transformer-based algorithms perform better on longer ***,a brief comparison of all these techniques is provided,and it concluded that transformers have the potential to revolutionize Natural Language Processing(NLP)methods in the near future.

关键词： Supervised learning algorithms fake news classification online disinformation transformers recurrent neural network(RNN)disinformation transformers recurrent neural network(RNN)

来源：评论

学校读者我要写书评

暂无评论

Environment-Tolerant Trust Opportunity Routing based on Reinforcement Learning for Internet of Underwater Things

引用

IEEE Transactions on Mobile Computing 2025年第07期24卷 6348-6360页

作者： He, Yu Han, Guangjie Hou, Yun Lin, Chuan Hohai University College of Information Science and Engineering Changzhou213200 China Hohai University Key Laboratory of Maritime Intelligent Network Information Technology Ministry of Education Changzhou213200 China Hohai University College of Computer Science and Software Engineering Nanjing211100 China Northeastern University Software College Shenyang110819 China

The Internet of Underwater Things (IoUT) has garnered significant interest due to its potential applications in monitoring underwater environments. However, the unique characteristics of acoustic communication, such as long propagation delays and high attenuation, present considerable obstacles for achieving efficient and dependable data transmission. Opportunistic routing is a crucial technique for enhancing packet delivery ratios by selecting a set of forwarding nodes and utilizing their cooperative forwarding to boost network throughput. Nevertheless, choosing an excessive number of forwarding nodes can lead to wasteful energy usage and extended communication delays. Moreover, the overlooked trustworthiness of forwarded nodes in most research works can undermine the effectiveness of opportunistic routing. Therefore, this study presents a novel trust opportunistic routing scheme that employs reinforcement learning to achieve resilience in constantly changing underwater settings. The combination of reinforcement learning and trust management enables the proposed opportunistic routing scheme to adapt to the unstable underwater environment and unknown malicious attacks. Initially, a method is introduced for measuring environmental fitness by considering multiple trust factors, including communication success rate, data reliability, and location dynamics. The proposed scheme then uses reinforcement learning to develop a reliable opportunistic routing method based on quantified state information. This component employs the obtained state to formulate action strategies and obtains reward values from environmental inputs. The reward update equation integrates these qualities to optimize the deployment of superior action strategies, finally achieving trust opportunistic routing for underwater data collection. Fundamental experimental results demonstrate that the proposed protocol performs exceptionally well in demanding underwater conditions, outperforming existing method

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Application of Lightweight DSC-NET Network Model in Histopathological Images of Lung Cancer and Colon Cancer

Application of Lightweight DSC-NET Network Model in Histopat...

引用

2023 International Conference on Cognitive Computing and Complex Data, ICCD 2023

作者： An, Wenlong Zong, Hui Zhou, Heng Wang, Jian Huaiyin Institute of Technology Faculty of Computer and Software Engineering Huaian China

ISBN: (纸本)9798350306293

Classification of histopathological images is a fundamental task in the workflow of pathological diagnosis. Due to the complexity of pathological images, it is particularly important to use deep learning to improve diagnostic efficiency. This paper designs a lightweight network model DSC-NET, which consists of multi-scale feature stitching Multi-Conv, coordinate attention CA. The improved selfcalibrated convolution MSC-Conv is composed of selfcalibrated convolution SC-Conv, coordinate attention CA and depthwise DW. In addition, the DSC-NET network model converts the 1×1 convolutional layer (Conv) in the Block module into a linear layer (Linear), which reduces the computational complexity of the model while maintaining the ability of the convolution operation to capture local features. The research in this paper adopts the lung cancer and colon cancer datasets and adds Gaussian noise to these datasets to simulate the equipment shooting situation and evaluate the lightweight DSC-NET network model. Through quantitative comparisons with previous state-of-the-art methods, our experimental results demonstrate that the proposed method achieves superior accuracy. Furthermore, our method stands out with a smaller parameter count and significantly lower FLOPs, highlighting its efficiency and computational advantages. It has important potential to assist pathologists in pathological diagnosis. © 2023 IEEE.

关键词： Convolution

来源：评论

学校读者我要写书评

暂无评论

Automatic Anomaly Detection from IoT-Time Series Dataset and Evaluation of Performance Metrics 23th

Automatic Anomaly Detection from IoT-Time Series Dataset and...

引用

23rd International Conference on Intelligent Systems Design and Applications, ISDA 2023

作者： Sunitha, Lingam Srinivas, B. Sunil Ramasahayam, Shravya Department of Information Technology Vardhaman College of Engineering Hyderabad India TKR College of Engineering and Technology Hyderabad India Software Development Engineer Flipkart Bangalore India

ISBN: (纸本)9783031647758

In the current era, anomaly detection (AD) has become an important area of research. Considering that the majority of prior strategies needed domain – specific supervision to set the model parameters and that the presently used techniques only support categorical or numerical data. Consequently, there exists a necessity. For contemporary algorithms that are independent of a domain provides generalized methods for datasets with mixed attributes, as well. The proposed approach can automatically differentiate between anomalies and normal behavior when employed with data that have numeric or categorical, i.e., single data type along with data with mixed-type attributes. The main objective of the suggested work is to improve classification performance by automatically removing abnormalities. The current study employs Score Based Anomaly Detection (SBAD), a Gaussian Mixture Model-based anomaly detection technique. Buildings can record environmental data using small computers like Arduino devices, which can then be used to anticipate basic and helpful qualities. Data from IoT-based time series is gathered from the UCI repository. The data set includes the date as well as environmental parameters including light, temperature, humidity, and CO2. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2024.

关键词： Anomaly detection

来源：评论

学校读者我要写书评

暂无评论

LLMaaS: Serving Large-Language Models on Trusted Serverless Computing Platforms

IEEE Transactions on Artificial Intelligence

引用

IEEE Transactions on Artificial Intelligence 2025年第2期6卷 405-415页

作者： Cai, Zinuo Ma, Rongbo Fu, Yicheng Zhang, Weishan Ma, Ruhui Guan, Haibing Shanghai Jiao Tong University School of Electronic Information and Electrical Engineering Shanghai200240 China Qingdao Institute of Software College of Computer Science and Technology Qingdao266555 China

In recent years, the emergence of large-language models (LLMs) has profoundly transformed our production and lifestyle. These models have shown tremendous potential in fields, such as natural language processing, speech recognition, and recommendation systems, and are increasingly playing crucial roles in applications such as human-computer interaction and intelligent customer service. Efficient inference solutions for LLMs in data centers have been extensively researched, with a focus on meeting users' quality of service requirements. In this article, we focus on two additional requirements that responsible LLM inference should meet under QoS conditions: security throughout the model execution process and low maintenance requirements for the inference system. Therefore, we propose LLMaaS, a trusted model inference platform based on a serverless computing platform aimed at providing inference as a service for LLMs. First, we design a trusted serverless computing platform based on software guard extension (SGX), which includes distributed identity verification and SGX device plugins to ensure the security and trustworthiness of the inference process. Additionally, to reduce the maintenance requirements of the system, we enhance the SGX-based deep learning computing framework, including replacing PyTorch and using a greedy algorithm for graph partitioning. We conduct tests on four typical large models, and the experimental results demonstrate that, with minimal overhead and user code modifications, we can ensure the security of model execution. © 2020 IEEE.

关键词： Quality of service

来源：评论

学校读者我要写书评

暂无评论

A Multi-constraint 3D Mesh Models Deformation Method

A Multi-constraint 3D Mesh Models Deformation Method

引用

2023 International Conference on Cognitive Computing and Complex Data, ICCD 2023

作者： Zhang, Chuangchuang Chen, Xiaobing Lin, Jianchu Guo, Shuxin Wang, Rongrong Gu, Yinxi Huaiyin Insititute of Technology Faculty of Computer and Software Engineering Huai'an China

ISBN: (纸本)9798350306293

Aiming at the problems of too many control vertices and difficult operation of the traditional free deformation technique, a multi-constraint 3D mesh models deformation method is proposed. Firstly, the input model is embedded into a sparse tetrahedral control mesh, and the positional relationship between the model and the control mesh is established using tetrahedral coordinates. Secondly, skeleton nodes are set on the model to generate skeleton segments, control mesh vertices are bound to the skeleton segments, the control mesh is deformed by manipulating the rotation and translation of the skeleton nodes, and finally the new positions of the model vertices are computed based on the control mesh. Since the skeleton drives a sparse tetrahedral control mesh, the accuracy of model deformation will be reduced, resulting in a series of problems in the deformed model. In order to optimize the model deformation, this paper sets the influence range of the control mesh on the model deformation by the weight of the constraint region, adds point constraints to the control mesh vertices to set the change range of this vertex, and finally calculates the correction factor according to the stretch constraints and volume constraints, and optimally adjusts the position of the control meshes vertices. The experimental results show that this method can make up for the problems of the traditional free deformation technique, which is complicated and unintuitive to operate, and improve the accuracy of deforming the model. © 2023 IEEE.

关键词： Mesh generation

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：