检索结果-内蒙古大学图书馆

SSRN 2024年

作者： Zhong, Xiaojing Su, Yukun Wu, Zhonghua Lin, Guosheng Wu, Qingyao School of Software Engineering South China University of Technology China School of Computer Science and Engineering Nanyang Technological University Singapore Key Laboratory of Big Data and Intelligent Robot Ministry of Education China Peng Cheng Laboratory China Tencent Wechat Department China SenseTime Research

3D virtual try-on tasks aim to generate realistic try-on results for full- body garments, allowing them to be observed from arbitrary perspectives. Recent methods often represent the 3D human form with a fixed topol- ogy, limiting garment representation and leading to higher computational demands. Furthermore, the challenge of flexible transfer over various body parts remains. To address these problems, we introduce DI-Net(Decomposed Implicit Garment Transfer Network), where garment transfer is decomposed into 2D image layout transfer and 3D texture layout transfer. We propose a complementary warping module that leverages dense correspondence learn- ing and sparse flow learning to align the body appearance and garments with the target pose for image layout transfer. Subsequently, the transferred garments are lifted to the 3D human mesh by constructing a pixel-aligned implicit function and predicting the texture based on geometric priors. By applying different semantic masks for garment transfer, we efficiently control the texture of the generated mesh. Experimental results show the effective- ness and superiority of our method in the 3D virtual try-on task. © 2024, The Authors. All rights reserved.

关键词： Textures

来源：评论

学校读者我要写书评

暂无评论

Once read is enough: domain-specific pretraining-free language models with cluster-guided sparse experts for long-tail domain knowledge 24

Once read is enough: domain-specific pretraining-free langua...

引用

Proceedings of the 38th International Conference on Neural Information Processing Systems

作者： Fang Dong Mengyi Chen Jixian Zhou Yubin Shi Yixuan Chen Mingzhi Dong Yujiang Wang Dongsheng Li Xiaochen Yang Rui Zhu Robert Dick Qin Lv Fan Yang Tun Lu Ning Gu Li Shang China and Shanghai Key Laboratory of Data Science School of Computer Science Fudan University Oxford Suzhou Centre for Advanced Research Suzhou China University of Bath and China and Shanghai Key Laboratory of Data Science School of Computer Science Fudan University Microsoft Research Asia Shanghai China School of Mathematics & Statistics University of Glasgow Bayes Business School City St George's University of London Department of Electrical Engineering and Computer Science University of Michigan Department of Computer Science University of Colorado Boulder School of Microelectronics Fudan University

ISBN: (纸本)9798331314385

Language models (LMs) only pretrained on a general and massive corpus usually cannot attain satisfying performance on domain-specific downstream tasks, and hence, applying domain-specific pretraining to LMs is a common and indispensable practice. However, domain-specific pretraining can be costly and time-consuming, hindering LMs' deployment in real-world applications. In this work, we consider the incapability to memorize domain-specific knowledge embedded in the general corpus with rare occurrences and "long-tail" distributions as the leading cause for pretrained LMs' inferior downstream performance. Analysis of Neural Tangent Kernels (NTKs) reveals that those long-tail data are commonly overlooked in the model's gradient updates and, consequently, are not effectively memorized, leading to poor domain-specific downstream performance. Based on the intuition that data with similar semantic meaning are closer in the embedding space, we devise a Cluster-guided Sparse Expert (CSE) layer to actively learn long-tail domain knowledge typically neglected in previous pretrained LMs. During pretraining, a CSE layer efficiently clusters domain knowledge together and assigns long-tail knowledge to designate extra experts. CSE is also a lightweight structure that only needs to be incorporated in several deep layers. With our training strategy, we found that during pretraining, data of long-tail knowledge gradually formulate isolated, "outlier" clusters in an LM's representation spaces, especially in deeper layers. Our experimental results show that only pretraining CSE-based LMs is enough to achieve superior performance than regularly pretrained-finetuned LMs on various downstream tasks, implying the prospects of domain-specific-pretraining-free language models.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Temporal Feature Matters: A Framework for Diffusion Model Quantization

arXiv

引用

arXiv 2024年

作者： Huang, Yushi Gong, Ruihao Liu, Xianglong Liu, Jing Li, Yuhang Lu, Jiwen Tao, Dacheng Electrical and Computer Engineering Department Hong Kong University of Science and Technology Hong Kong State Key Laboratory of Software Development Environment Beihang University China Department of Data Science and AI Faculty of IT Monash University Australia Electrical Engineering Department Yale University United States Department of Automation Tsinghua University China School of Computer Science and Engineering Nanyang Technological University Singapore

The Diffusion models, widely used for image generation, face significant challenges related to their broad applicability due to prolonged inference times and high memory demands. Efficient Post-Training Quantization (PTQ) is crucial to address these issues. However, unlike traditional models, diffusion models critically rely on the time-step for the multi-round denoising. Typically, each time-step is encoded into a hypersensitive temporal feature by several modules. Despite this, existing PTQ methods do not optimize these modules individually. Instead, they employ unsuitable reconstruction objectives and complex calibration methods, leading to significant disturbances in the temporal feature and denoising trajectory, as well as reduced compression efficiency. To address these challenges, we introduce a novel quantization framework that includes three strategies: 1) TIB-based Maintenance: Based on our innovative Temporal Information Block (TIB) definition, Temporal Information-aware Reconstruction (TIAR) and Finite Set Calibration (FSC) are developed to efficiently align original temporal features. 2) Cache-based Maintenance: Instead of indirect and complex optimization for the related modules, pre-computing and caching quantized counterparts of temporal features are developed to minimize errors. 3) Disturbance-aware Selection: Employ temporal feature errors to guide a fine-grained selection between the two maintenance strategies for further disturbance reduction. This framework preserves most of the temporal information and ensures high-quality end-to-end generation. Extensive testing on various datasets, diffusion models and hardware confirms our superior performance and acceleration. © 2024, CC BY.

关键词： Spatio-temporal data

来源：评论

学校读者我要写书评

暂无评论

LF-YOLOv7: Improved YOLOv7 Based on Lightweight Modules and Novel Feature Fusion for Object Detection on Drone-Captured Scenarios

LF-YOLOv7: Improved YOLOv7 Based on Lightweight Modules and ...

引用

International Conference on Computational Science and Computational Intelligence (CSCI)

作者： Wangyu Jiang Le Wang Guojun Mao Meng Sun Fayaz Ali Dharejo Ghulam Ali Mallah Fujian Provincial Key Lab of Big Data Mining&Applications Fujian University of Technology Fuzhou Fujian Province China Department of Electrical Engineering and Computer Science Khalifa Univeristy AbuDhabi UAE Department of Computer Science Shah Abdul Latif University Khairpur Pakistan

ISBN: (数字)9798350361513

ISBN: (纸本)9798350372304

To address challenges in drone-captured images, including small-sized objects, multiple scales, and object diversity, we propose a novel lightweight and feature fusion based object detection model, called LF-YOLOv7. Firstly, a comprehensive high-resolution refinement design is investigated through simultaneously restructuring the head and optimizing the backbone, achieving more accurate assignment of multiple anchor boxes of different sizes and higher spatial resolution with lightweight mode. Furthermore, we present a new feature fusion module SM-BiFPN, which effectively integrates shallow details through cross-layer connections and enables the model to obtain more feature information related to small objects. Experimental evaluation shows that our method achieves a more competitive results than the benchmark and other compared state-of-the-art models in terms of mAP50 and the number of parameters.

关键词： Cross layer design Head Accuracy Scientific computing Computational modeling Object detection Benchmark testing

来源：评论

学校读者我要写书评

暂无评论

Energy-efficient and Privacy-preserving Edge Intelligence for 6G-Empowered Intelligent Transportation Systems

引用

IEEE Network 2025年

作者： Zhu, Chenlu Fan, Xiaoxuan Deng, Xianjun Xiao, Ziheng Guo, Tingting Liu, Shenghao Sun, Jiaqi Park, Jong Hyuk Hubei Key Laboratory of Distributed System Security Hubei Engineering Research Center on Big Data Security School of Cyber Science and Engineering Huazhong University of Science and Technology Wuhan430074 China Wuhan Kedao Geographic Information Engineering Co. Wuhan China Department of Computer Science and Engineering Seoul National University of Science and Technology Seoul Korea Republic of

6G-empowered Intelligent Transportation Systems (ITS) generate large amounts of data through millions of devices and sensors at the terminal and network edge. Edge intelligence advances the frontier of data-driven artificial intelligence (AI) to the network edge, which enables AI applications to better utilize edge data. Broader connectivity, higher device density, and stronger data exchange present challenges for reliable device communication and privacy data protection in 6G-empowered ITS. To address these challenges, this paper proposes an energy-efficient and privacy-preserving edge intelligence framework named EP-EI. EP-EI deploys a deep reinforcement learning-based cluster head selection model at the edge to ensure energy-efficient communication for terminal devices. Meanwhile, EP-EI designs a privacy-aware hierarchical computing framework for terminal-edge-cloud to prevent privacy leakage. Furthermore, the application scenarios, challenges, and future directions of EP-EI are discussed. Finally, simulation experiments in a specific intelligent transportation case study demonstrate that EP-EI can provide energy-efficient and privacy-protecting services. © 2025 IEEE.

关键词： Differential privacy

来源：评论

学校读者我要写书评

暂无评论

GraphBEV: Towards Robust BEV Feature Alignment for Multi-Modal 3D Object Detection

arXiv

引用

arXiv 2024年

作者： Song, Ziying Yang, Lei Xu, Shaoqing Liu, Lin Xu, Dongyang Jia, Caiyan Jia, Feiyang Wang, Li School of Computer and Information Technology Beijing Jiaotong University China Beijing Key Lab of Traffic Data Analysis and Mining China School of Vehicle and Mobility Tsinghua University China Department of Electrome chanical Engineering University of Macau China School of Mechanical Engineering Beijing Institute of Technology China

Integrating LiDAR and camera information into Bird’s-Eye-View (BEV) representation has emerged as a crucial aspect of 3D object detection in autonomous driving. However, existing methods are susceptible to the inaccurate calibration relationship between LiDAR and the camera sensor. Such inaccuracies result in errors in depth estimation for the camera branch, ultimately causing misalignment between LiDAR and camera BEV features. In this work, we propose a robust fusion framework called GraphBEV. Addressing errors caused by inaccurate point cloud projection, we introduce a LocalAlign module that employs neighbor-aware depth features via Graph matching. Additionally, we propose a GlobalAlign module to rectify the misalignment between LiDAR and camera BEV features. Our GraphBEV framework achieves state-of-the-art performance, with an mAP of 70.1%, surpassing BEVFusion by 1.6% on the nuScenes validation set. Importantly, our GraphBEV outperforms BEVFusion by 8.3% under conditions with misalignment noise. Copyright © 2024, The Authors. All rights reserved.

关键词： Object detection

来源：评论

学校读者我要写书评

暂无评论

Image Forgery Localization with State Space Models

arXiv

引用

arXiv 2024年

作者： Lou, Zijie Cao, Gang Guo, Kun Weng, Shaowei Yu, Lifang School of Computer and Cyber Sciences Communication University of China Beijing100024 China State Key Laboratory of Media Convergence and Communication Communication University of China Beijing100024 China Fujian Provincial Key Laboratory of Big Data Mining and Applications Fujian University of Technology Fuzhou350118 China Department of Information Engineering Beijing Institute of Graphic Communication Beijing100026 China

Pixel dependency modeling from tampered images is pivotal for image forgery localization. Current approaches predominantly rely on Convolutional Neural Networks (CNNs) or Transformer-based models, which often either lack sufficient receptive fields or entail significant computational overheads. Recently, State Space Models (SSMs), exemplified by Mamba, have emerged as a promising approach. They not only excel in modeling long-range interactions but also maintain a linear computational complexity. In this paper, we propose LoMa, a novel image forgery localization method that leverages the selective SSMs. Specifically, LoMa initially employs atrous selective scan to traverse the spatial domain and convert the tampered image into ordered patch sequences, and subsequently applies multi-directional state space modeling. In addition, an auxiliary convolutional branch is introduced to enhance local feature extraction. Extensive experimental results validate the superiority of LoMa over CNN-based and Transformer-based state-of-the-arts. To our best knowledge, this is the first image forgery localization model constructed based on the SSM-based model. We aim to establish a baseline and provide valuable insights for the future development of more efficient and effective SSM-based forgery localization models. Code is available at https://***/multimediaFor/LoMa. Copyright © 2024, The Authors. All rights reserved.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Evaluation of China's Digital Economy: A Case Study using Entropy and TOPSIS 11

Evaluation of China's Digital Economy: A Case Study using En...

引用

11th International Conference on Information Technology and Quantitative Management, ITQM 2024

作者： Liu, Junjie Wang, Boya Xue, Jiayu Qu, Yi Shi, Yong Sino-Danish College University of Chinese Academy of Sciences Beijing100049 China Research Center on Fictitious Economy and Data Science Chinese Academy of Sciences Beijing100190 China Key Laboratory of Big Data Mining and Knowledge Management Chinese Academy of Sciences Beijing100190 China Faculty of Science and Engineering University of Nottingham Ningbo China Zhejiang Ningbo315100 China School of Computer Science and Technology University of Chinese Academy of Sciences Beijing101408 China School of Economics and Management University of Chinese Academy of Sciences Beijing100190 China College of Information Science and Technology University of Nebraska at Omaha OmahaNE68182 United States

Recently, digital economy in social science has become increasingly significant. A well-constructed evaluation structure could generally and promptly describe and reveal its overall social digitalization development. Building on existent researches, this paper improves the evaluation system, which comprises four first-level indicators and thirty-one second-level indicators. Notably, this paper innovatively incorporates higher education into digital innovation of the evaluation. The Entropy method is utilized to assign weights to various indicators in an objective way, with employing the TOPSIS approach to prioritize the development of digital economy in places based on the closeness to the ideal solution. By analysing related data related from 31 provinces in China between 2017 to 2021, this paper greatly finds: Firstly, the development of digital economy vary significantly across provinces, with most provinces at a relatively poor level, indicating considerable potential for overall growth;Secondly, Guangdong, Jiangsu, Zhejiang, and Shandong consistently ranked in the top four from 2017 to 2021, reflecting their specific development attributes and their emphasis on digital economy;Thirdly, significant disparities exist both between regions and among provinces. Consequently, this paper suggests establishing a comprehensive national digital economy development plan, formulating region-specific digital economy policies tailored to local conditions, and encouraging leading provinces to actively share experiences with other provinces. © 2024 The Authors.

关键词： Digital economy Evaluation Multi-source information TOPSIS

来源：评论

学校读者我要写书评

暂无评论

Progress in research on ultrasound radiomics for predicting the prognosis of breast cancer

引用

Cancer Innovation 2023年第4期2卷 283-289页

作者： Xuantong Gong Xuefeng Liu Xiaozheng Xie Yong Wang Department of Ultrasound National Cancer Center/National Clinical Research Center for Cancer/Cancer HospitalChinese Academy of Medical Sciences and Peking Union Medical CollegeBijingChina State Key Laboratory of Virtual Reality Technology and Systems School of Computer Science and EngineeringBejjing Advanced Innovation Center for Big Data and Brain Computing(BDBC)Beihang UniversityBeijingChina School of Computer and Communication Engi neering.University of Science and Technology Beijing BeijingChina

Breast cancer is the most common malignant tumor and the leading cause of cancer-related deaths in women *** means of predicting the prognosis of breast cancer are very helpful in guiding treatment and improving patients'*** extracted by radiomics reflect the genetic and molecular characteristics of a tumor and are related to its biological behavior and the patient's ***,radiomics provides a new approach to noninvasive assessment of breast cancer *** is one of the commonest clinical means of examining breast *** recent years,some results of research into ultrasound radiomics for diagnosing breast cancer,predicting lymph node status,treatment response,recurrence and survival times,and other aspects,have been *** this article,we review the current research status and technical challenges of ultrasound radiomics for predicting breast cancer *** aim to provide a reference for radiomics researchers,promote the development of ultrasound radiomics,and advance its clinical application.

关键词： breast cancer deeplearning prognosisprediction radiomics ultrasound

来源：评论

学校读者我要写书评

暂无评论

Snapshot boosting: a fast ensemble framework for deep neural networks

引用

Science China(Information Sciences) 2020年第1期63卷 77-88页

作者： Wentao ZHANG Jiawei JIANG Yingxia SHAO Bin CUI Center for Data Science Peking University National Engineering Laboratory for Big Data Analysis and Applications Department of Computer Science Beijing Key Lab of Intelligent Telecommunications Software and Multimedia School of Computer ScienceBeijing University of Posts and Telecommunications Key Lab of High Confidence Software Technologies Department of Computer SciencePeking University

Boosting has been proven to be effective in improving the generalization of machine learning models in many fields. It is capable of getting high-diversity base learners and getting an accurate ensemble model by combining a sufficient number of weak learners. However, it is rarely used in deep learning due to the high training budget of the neural network. Another method named snapshot ensemble can significantly reduce the training budget, but it is hard to balance the tradeoff between training costs and diversity. Inspired by the ideas of snapshot ensemble and boosting, we propose a method named snapshot boosting. A series of operations are performed to get many base models with high diversity and accuracy, such as the use of the validation set, the boosting-based training framework, and the effective ensemble strategy. Last, we evaluate our method on the computer vision(CV) and the natural language processing(NLP) tasks, and the results show that snapshot boosting can get a more balanced trade-off between training expenses and ensemble accuracy than other well-known ensemble methods.

关键词： ensemble learning deep learning boosting neural network snapshot ensemble

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：