检索结果-内蒙古大学图书馆

Conference on computer Vision and Pattern Recognition (CVPR)

作者： Yiduo Guo Bing Liu Dongyan Zhao Wangxuan Institute of Computer Technology Peking University Department of Computer Science University of Illinois Chicago BIGAI Beijing China National Key Laboratory of General Artificial Intelligence

Existing continual learning (CL) research regards catastrophic forgetting (CF) as almost the only challenge. This paper argues for another challenge in class-incremental learning (CIL), which we call cross-task class discrimination (CTCD), i.e., how to establish decision boundaries between the classes of the new task and old tasks with no (or limited) access to the old task data. CTCD is implicitly and partially dealt with by replay-based methods. A replay method saves a small amount of data (replay data) from previous tasks. When a batch of current task data arrives, the system jointly trains the new data and some sampled replay data. The replay data enables the system to partially learn the decision boundaries between the new classes and the old classes as the amount of the saved data is small. However, this paper argues that the replay approach also has a dynamic training bias issue which reduces the effectiveness of the replay data in solving the CTCD problem. A novel optimization objective with a gradient-based adaptive method is proposed to dynamically deal with the problem in the online CL process. Experimental results show that the new method achieves much better results in online CL.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Modeling the failures of power-aware data centers by leveraging heat recirculation

Modeling the failures of power-aware data centers by leverag...

引用

作者： Feng, Hao Deng, Yuhui Yu, Liang Department of Computer Science Jinan University Guangzhou China Key Laboratory of Computer Architecture Institute of Computing Technology Chinese Academy of Sciences Beijing China

With the explosive growth of data, hundreds of thousands of servers may be contained in a single data center. Hence, node failures are unavoidable and generally negatively effects the performance of the whole data center. Additionally, data centers with a large number of nodes will cause plenty of energy consumption. Many existing task scheduling techniques can effectively reduce the power consumption in data centers by considering heat recirculation. However, the traditional techniques do not take the situation of node failures into account. This paper proposes an airflow-based failure model for data centers by leveraging heat recirculation. In this model, the spatial distribution and time distribution of failures are considered. Furthermore, a genetic algorithm (GA) and a simulated annealing algorithm (SA) are implemented to evaluate the proposed failure model. Because the positions of node failures have a significant impact on the heat recirculation and the energy consumption of data centers, failures with different positions are analyzed and evaluated. The experimental results demonstrate that the energy consumption of data centers can be significantly reduced by using the GA and SA algorithms for task scheduling based on the proposed failure model. © 2019 John Wiley & Sons, Ltd.

关键词： Energy efficiency

来源：评论

学校读者我要写书评

暂无评论

DCGSD: Low-Light Image Enhancement with Dual-Conditional Guidance Sparse Diffusion Model

引用

IEEE Transactions on Circuits and Systems for Video Technology 2025年

作者： Jin, Haiyan Wang, Jing Zuo, Fengyuan Su, Haonan Xiao, Zhaolin Wang, Bin Zhang, Yuanlin Xi'an University of Technology Shaanxi Key Laboratory for Network Computing and Security Technology Department of Computer Science and Engineering NO. 5 South Jinhua Road Shaanxi Xi'an710048 China

When restoring low-light images, most methods largely overlook the ambiguity due to dark noise and lack discrimination for region and shape representations, resulting in invalid feature enhancement. In this work, we propose a physically explainable and prior guidance model for low-light image enhancement, termed Dual-Conditional Guidance Sparse Diffusion (DCGSD). Specifically, we introduce an elaborately designed Luminance Structure Guidance Head, which can be easily plugged into the existing diffusion model to emphasize the value of the luminance and structural representation. Furthermore, for reliable noise analysis, we provide a novel Sparse Attention Enhancement Module that is adaptively empowered to exploit the most useful region-to-region dependencies. This dynamic selection makes the diffusion process from dense to sparse, thus improving the efficiency of the reasoning noise distributions. To avoid noise amplification, we further present a Skip Calibration Module, which can be used to refine the local neighborhood that contains noisy and structural information. Extensive experiments have been performed to verify the superiority of the proposed method. DCGSD shows that leveraging dual-conditional guidance can support the diffusion model to produce sharper and more realistic results. © 1991-2012 IEEE.

关键词： Feature extraction

来源：评论

学校读者我要写书评

暂无评论

Runx2 controls the osteogenic fate of growth plate chondrocytes

引用

Genes & Diseases 2025年第3期12卷 1-3页

作者： Daofu Zeng Jiamin Yu Ke Lu Dan Yi Zhidao Xia Ling Qin Guozhi Xiao Xiao Yang Liping Tong Di Chen Research Center for Computer-aided Drug Discovery Shenzhen Institute of Advanced TechnologyChinese Academy of SciencesShenzhenGuangdong 518055China Faculty of Pharmaceutical Sciences Shenzhen University of Advanced TechnologyShenzhenGuangdong 518055China University of Chinese Academy of Sciences Chinese Academy of SciencesBeijing 100049China Institute of Life Science Swansea University Medical SchoolSwanseaSA28PPUK Musculoskeletal Research Laboratory of Department of Orthopaedics&Traumatology and Innovative Orthopaedic Biomaterial&Drug Translational Research Laboratory Li Ka Shing Institute of Health SciencesThe Chinese University of Hong KongHong Kong 518172China School of Medicine Southern University of Science and TechnologyShenzhenGuangdong 518055China State Key Laboratory of Proteomics Genetic Laboratory of Development and DiseasesInstitute of BiotechnologyBeijing 100071China

The origin of bone marrow osteoblasts is not totally *** findings demonstrated that bone marrow osteoblasts could be derived from a subpopulation of hypertrophic Col2^(+)/Col10^(+)chondrocytes which migrate from the growth plate into the bone marrow cavity underneath the growth plate and dedifferentiate into mesenchymal progenitor cells and then differentiate into mature osteoblasts.1 This process is called chondrocyte-osteoblast *** type of osteoblast participates in bone formation and is involved in maintaining bone remodeling,especially in the epiphyseal and diaphyseal regions of long *** growth factors,such as Ihh,PTH,and Wnt signaling molecules have been demonstrated to play a critical role in the regulation of chondrocyte-osteoblast transdifferentiation2;however,the role of Runx2,the key transcription factor controlling skeletal development,3 in chondrocyte-osteoblast transdifferentiation has not been fully defined.

关键词： RUNX mesenchymal progenitor cells chondrocyte osteoblast transdifferentiation hypertrophic chondrocytes growth plate bone marrow bone marrow osteoblasts

来源：评论

学校读者我要写书评

暂无评论

LEARNING TO GENERATE STRUCTURED OUTPUT WITH SCHEMA REINFORCEMENT LEARNING

arXiv

引用

arXiv 2025年

作者： Lu, Yaxi Li, Haolun Cong, Xin Zhang, Zhong Wu, Yesai Lin, Yankai Liu, Zhiyuan Liu, Fangming Sun, Maosong Department of Computer Science and Technology Tsinghua University China Gaoling School of Artificial Intelligence Renmin University of China China Peng Cheng Laboratory China

This study investigates the structured generation capabilities of large language models (LLMs), focusing on producing valid JSON outputs against a given schema. Despite the widespread use of JSON in integrating language models with programs, there is a lack of comprehensive analysis and benchmarking of these capabilities. We explore various aspects of JSON generation, such as structure understanding, escaping, and natural language description, to determine how to assess and enable LLMs to generate valid responses. Building upon this, we propose SchemaBench features around 40K different JSON schemas to obtain and assess models’ abilities in generating valid JSON. We find that the latest LLMs are still struggling to generate a valid JSON string. Moreover, we demonstrate that incorporating reinforcement learning with a Fine-grained Schema Validator can further enhance models’ understanding of JSON schema, leading to improved performance. Our models demonstrate significant improvement in both generating JSON outputs and downstream tasks. Copyright © 2025, The Authors. All rights reserved.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Empowering Education: Harnessing Artificial Intelligence for Adaptive E-Learning Excellence

Empowering Education: Harnessing Artificial Intelligence for...

引用

International Conference on Artificial Intelligence and Its Practical Applications in the Age of Digital Transformation, I2COMSAPP 2024

作者： Benkhalfallah, Fateh Laouar, Mohamed Ridda Benkhalfallah, Mohamed Salah Tebessa Algeria Faculty of Exact Sciences and Natural and Life Sciences Echahid Cheikh Larbi Tebessi University Tebessa Algeria Artificial Intelligence and Autonomous Things Laboratory Oum El Bouaghi Algeria Department of Mathematics and Computer Science University of Oum El Bouaghi Oum El Bouaghi Algeria

ISBN: (纸本)9783031714283

The global landscape is currently undergoing an unprecedented and multifaceted qualitative and quantitative evolution, prominently exemplified in the realm of adaptive e-learning. This transformation is intricately linked to the pervasive integration of contemporary global technologies, particularly artificial intelligence (AI), which furnishes an alternative paradigm to conventional learning methodologies. The overarching objective is to elevate the efficacy and efficiency of learning endeavors while concurrently emphasizing sustainability and widespread accessibility. These advancements have precipitated radical shifts across all educational facets, notably in teaching and training methodologies, giving rise to novel mechanisms and strategies in knowledge and skill acquisition. The contemporary educational ethos seeks liberation from constraints of spatial, affiliative, and temporal limitations, aspiring to attain elevated proficiency levels across diverse dimensions. This paper aspires to elucidate and harness the latent potential of artificial intelligence within the context of adaptive e-learning excellence. It achieves this goal through a meticulous review and comparative analysis of pertinent literature studies, delineating the salient similarities, differences, and key features characterizing major types of intelligent e-learning. Additionally, it delineates e-learning patterns and provides an insightful description of the primary formats for the application of artificial intelligence in education, articulates AI-powered adaptive learning models, and underscores emerging trends within the expansive intersection of AI and the educational domain. Furthermore, it engages in a comprehensive discussion of the potential challenges, limitations, and practical considerations of implementing AI in e-learning. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2024.

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

Speaker Diarization in Overlapped Speech 19

Speaker Diarization in Overlapped Speech

引用

19th IEEE International Multi-Conference on Systems, Signals and Devices, SSD 2022

作者： Bounazou, Hadjer Asbai, Nassim Zitouni, Sihem Mohamed Ei Bachir Ei Ibrahimi University Eta Laboratory Department of Electronic. Bordj Bou Arreridj34030 Algeria Speech Com. & Signal Proc. Lab. Faculty of Electronics and Computer Sciences Usthb Bab Ezzouar16 111 Algeria Mohamed Ei Bachir Ei Ibrahimi University Faculty of Mathematics and Computer Science Department of Computer Science Bordj Bou Arreridj34030 Algeria

ISBN: (数字)9781665471084

ISBN: (纸本)9781665471084

In this paper, the effect of the overlapping speech on speaker diarization performances developed in our work, using NIST evaluation metrics (DER: diarization error rate, MR: misclassification rate and ACP: average cluster purity) on NIT2005 database is studied. Using cepstral features MFCCs for a speaker diarization system, its evaluation is done by including the overlapped speech regions. If an overlapped speech region is assigned to only one speaker, this region is considered as Missed speech for the second speaker. Through our experiences, the effects of this strategy using various overlapping speech scenarios show that the speaker diarization performances (in terms of DER, MR and ACP) are drastically degraded in all situations (DER=21.08%, MR=23.83% and ACP=80.14). © 2022 IEEE.

关键词： Performance evaluation Error analysis Databases Cepstral analysis NIST Task analysis

来源：评论

学校读者我要写书评

暂无评论

Close contact behaviors between medical staff and patients in dental clinics 18

Close contact behaviors between medical staff and patients i...

引用

18th International Conference on Indoor Air Quality and Climate, INDOOR AIR 2024

作者： Luo, Yingjie Zhao, Fangli Cao, Bing Dou, Zhiyang Zhang, Nan Xiao, Shenglan Sun Yat-sen University Shenzhen Campus Shenzhen China Sun Yat-sen University Guangzhou China Beijing Key Laboratory of Green Built Environment and Energy Efficient Technology Beijing University of Technology Beijing China Department of Computer Science The University of Hong Kong Hong Kong

ISBN: (纸本)9798331306816

Obtaining real-world data on close contact behaviors between dental staff and patients in clinics is of great significance for a comprehensive analysis of the pathogen transmission mechanism in this environment. We collected images of close contact behaviors between dental staff and patients during the diagnosis and treatment processes. The data were processed and identified through semi-supervision learning. Our findings revealed a relatively high incidence of close contact among personnel within dental clinics. Moreover, compared with nurses, the interpersonal distance between doctors and patients was significantly lower (P © 2024 18th Conference of the International Society of Indoor Air Quality and Climate, INDOOR AIR 2024 - Conference Program and Proceedings. All rights reserved.

关键词： Indoor air pollution

来源：评论

学校读者我要写书评

暂无评论

Effective Entry-Wise Flow for Molecule Generation 40

Effective Entry-Wise Flow for Molecule Generation

引用

40th IEEE International Conference on Data Engineering, ICDE 2024

作者： Zhang, Qifan Yao, Junjie Yang, Yuquan Shi, Yizhou Gao, Wei Wang, Xiaoling School of Computer Science and Technology East China Normal University China Software Engineering Institute East China Normal University China Shanghai Key Laboratory of Trustworthy Computing China University of Massachusetts Department of Chemistry Boston United States University of Michigan Ann Arbor United States

ISBN: (纸本)9798350317152

Molecule generation is a critical process in the fields of drug discovery and materials science. Recently, generative models based on normalizing flows have demonstrated significant potential in this domain. These models are particularly suited for handling the symmetrical and complex chemical structures often encountered in molecular datasets. Despite their promising nature, normalizing flow-based models for molecule generation face considerable challenges. The complexity of molecule representation, the rigorous demands of optimization, and the scarcity of training labels in molecular datasets contribute to these difficulties. Additionally, adequately and comprehensively learning the distribution of molecular datasets remains a formidable task. In this paper, we delve into the intricate entry-wise modules in vanilla flows, introducing an effective variation of flow-based models. Our proposed approach innovatively encapsulates affine coupling transformations within normalizing flows. Furthermore, we deconstruct existing invertible flow models, integrating them with newly developed entry-wise transformations. Our experimental studies demonstrate that these proposed entry-wise modules, when incorporated into standard flow-based models, surpass other generative models in performance on various representative datasets and generation tasks. Notably, in the context of low-resourced molecular graph generation, our model achieves remarkable performance compared to its counterparts. © 2024 IEEE.

关键词： Complex networks

来源：评论

学校读者我要写书评

暂无评论

Hybrid near- and far-field three-stage beam training with beam split for RIS-assisted OFDM communications

引用

Frontiers of Information Technology & Electronic Engineering 2024年第12期25卷 1664-1678页

作者： Zhichao CHENG Shupei ZHANG Shu FU Boya DI School of Electronics Engineering and Computer Science Peking UniversityBeijing 100871China State Key Laboratory of Advanced Optical Communication Systems and Networks School of ElectronicsPeking UniversityBeijing 100871China Department of Microelectronics and Communication Engineering Chongqing UniversityChongqing 400044China

With the development of millimeter-wave(mmWave)communication systems,large-scale reconfigurable intelligent surfaces(RISs)have gained considerable attention as a promising technology for signal strength enhancement and coverage ***,as the antenna scale and bandwidth increase,RIS-assisted wideband orthogonal frequency division multiplexing(OFDM)communication systems face challenges due to the near-field range expansion and the beam split effect over the high-frequency band,complicating the acquisition of channel state information(CSI).To tackle these challenges,we present a codebook-based three-stage beam training scheme by using the beam split effect to bypass CSI ***,by analyzing the beam split effect in RIS-assisted OFDM communication systems,we propose a beam-split-aware codebook capable of covering both the near and far fields with fewer codewords compared to conventional narrow-band *** such a codebook,a three-stage beam training mechanism is adopted to obtain the optimal codeword with low time overhead,thereby facilitating subsequent *** results demonstrate that the proposed scheme outperforms existing near-and far-field codebook-based schemes in terms of the beam training resolution and sum rate in the hybrid near–far field.

关键词： Near-far field Beam split Codebook design Beam training Reconfigurable intelligent surfaces

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：