检索结果-内蒙古大学图书馆

arXiv 2023年

作者： Xu, Fangzhi Liu, Jun Lin, Qika Zhao, Tianzhe Zhang, Jian Zhang, Lingling School of Computer Science and Technology Xi’an Jiaotong University Xi’an China Shaanxi Province Key Laboratory of Satellite and Terrestrial Network Tech. R&D National Engineering lab for Big Data Analytics Xi’an China

Logical reasoning task involves diverse types of complex reasoning over text, based on the form of multiple-choice question answering. Given the context, question and a set of options as the input, previous methods achieve superior performances on the full-data setting. However, the current benchmark dataset has the ideal assumption that the reasoning type distribution on the train split is close to the test split, which is inconsistent with many real application scenarios. To address it, there remain two problems to be studied: (1) How is the zero-shot capability of the models (train on seen types and test on unseen types)? (2) How to enhance the perception of reasoning types for the models? For problem 1, we propose a new benchmark for generalized zero-shot logical reasoning, named ZsLR. It includes six splits based on the three type sampling strategies. For problem 2, a type-aware model TaCo is proposed. It utilizes both the heuristic input reconstruction and the contrastive learning to improve the type perception in the global representation. Extensive experiments on both the zero-shot and full-data settings prove the superiority of TaCo over the state-of-the-art methods. Also, we experiment and verify the generalization capability of TaCo on other logical reasoning dataset. © 2023, CC BY.

关键词： Benchmarking

来源：评论

学校读者我要写书评

暂无评论

SAMPLING LOVÁSZ LOCAL LEMMA FOR GENERAL CONSTRAINT SATISFACTION SOLUTIONS IN NEAR-LINEAR TIME

arXiv

引用

arXiv 2022年

作者： He, Kun Wang, Chunyang Yin, Yitong The Key Lab of Data Engineering and Knowledge Engineering MOE Renmin University of China No. 59 Zhongguancun Street Haidian District Beijing China State Key Laboratory for Novel Software Technology Nanjing University 163 Xianlin Avenue Jiangsu Province Nanjing China

We give a fast algorithm for sampling uniform solutions of general constraint satisfaction problems (CSPs) in a local lemma regime. Suppose that the CSP has n variables with domain size at most q, each constraint contains at most k variables, shares variables with at most Δ constraints, and is violated with probability at most p by a uniform random assignment. The algorithm returns an almost uniform satisfying assignment in expected poly(q,k, Δ) ·(n) time, as long as a local lemma condition is satisfied: k · p · q2 · Δ5 ≤ C0 for a suitably small absolute constant C0. Previously, under similar local lemma conditions, sampling algorithms with running time polynomial in both nand Δ were only known for the almost atomic case, where each constraint is violated by a small number of forbidden local configurations. The key term Δ5 in our local lemma condition also improves the previously best known Δ7 for general CSPs [JPV21b] and Δ5.714 for atomic CSPs, including the special case of k-CNF [JPV21a, HSW21]. Our sampling approach departs from previous fast algorithms for sampling LLL, which were based on Markov chains. A crucial step of our algorithm is a recursive marginal sampler that is of independent interests. Within a local lemma regime, this marginal sampler can draw a random value for a variable according to its marginal distribution, at a cost independent of the size of the CSP. Copyright © 2022, The Authors. All rights reserved.

关键词： Constraint satisfaction problems

来源：评论

学校读者我要写书评

暂无评论

DISENTANGLING THE PROSODY AND SEMANTIC INFORMATION WITH PRE-TRAINED MODEL FOR IN-CONTEXT LEARNING BASED ZERO-SHOT VOICE CONVERSION

arXiv

引用

arXiv 2024年

作者： Chen, Zhengyang Wang, Shuai Zhang, Mingyang Liu, Xuechen Yamagishi, Junichi Qian, Yanmin Auditory Cognition and Computational Acoustics Lab MoE Key Lab of Artificial Intelligence AI Institute Department of Computer Science and Engineering Shanghai Jiao Tong University Shanghai China Shenzhen Research Institute of Big Data Shenzhen China Chinese University of Hong Kong Shenzhen China National Institute of Informatics Tokyo Japan

Voice conversion (VC) aims to modify the speaker's timbre while retaining speech content. Previous approaches have tokenized the outputs from self-supervised into semantic tokens, facilitating disentanglement of speech content information. Recently, in-context learning (ICL) has emerged in text-to-speech (TTS) systems for effectively modeling specific characteristics such as timbre through context conditioning. This paper proposes an ICL capability enhanced VC system (ICL-VC) employing a mask and reconstruction training strategy based on flow-matching generative models. Augmented with semantic tokens, our experiments on the LibriTTS dataset demonstrate that ICL-VC improves speaker similarity. Additionally, we find that k-means is a versatile tokenization method applicable to various pre-trained models. However, the ICL-VC system faces challenges in preserving the prosody of the source speech. To mitigate this issue, we propose incorporating prosody embeddings extracted from a pre-trained emotion recognition model into our system. Integration of prosody embeddings notably enhances the system's capability to preserve source speech prosody, as validated on the Emotional Speech database. Copyright © 2024, The Authors. All rights reserved.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

Disentangling The Prosody And Semantic Information With Pre-Trained Model For In-Context Learning Based Zero-Shot Voice Conversion

Disentangling The Prosody And Semantic Information With Pre-...

引用

IEEE Spoken Language Technology Workshop

作者： Zhengyang Chen Shuai Wang Mingyang Zhang Xuechen Liu Junichi Yamagishi Yanmin Qian Department of Computer Science and Engineering Auditory Cognition and Computational Acoustics Lab MoE Key Lab of Artificial Intelligence AI Institute Shanghai Jiao Tong University Shanghai China Shenzhen Research Institute of Big Data Shenzhen China Chinese University of Hong Kong Shenzhen China National Institute of Informatics Tokyo Japan

ISBN: (数字)9798350392258

ISBN: (纸本)9798350392265

Voice conversion (VC) aims to modify the speaker’s timbre while retaining speech content. Previous approaches have tokenized the outputs from self-supervised into semantic tokens, facilitating disentanglement of speech content information. Recently, in-context learning (ICL) has emerged in text-to-speech (TTS) systems for effectively modeling specific characteristics such as timbre through context conditioning. This paper proposes an ICL capability enhanced VC system (ICL-VC) employing a mask and reconstruction training strategy based on flow-matching generative models. Augmented with semantic tokens, our experiments on the LibriTTS dataset demonstrate that ICL-VC improves speaker similarity. Additionally, we find that k-means is a versatile tokenization method applicable to various pre-trained models. However, the ICL-VC system faces challenges in preserving the prosody of the source speech. To mitigate this issue, we propose incorporating prosody embeddings extracted from a pre-trained emotion recognition model into our system. Integration of prosody embeddings notably enhances the system’s capability to preserve source speech prosody, as validated on the Emotional Speech database.

关键词： Training Emotion recognition databases Semantics Self-supervised learning Speech enhancement Tokenization Text to speech Multilingual Timbre

来源：评论

学校读者我要写书评

暂无评论

FedPHE: A Secure and Efficient Federated Learning via Packed Homomorphic Encryption

引用

IEEE Transactions on Dependable and Secure Computing 2025年

作者： Li, Yuqing Yan, Nan Chen, Jing Wang, Xiong Hong, Jianan He, Kun Wang, Wei Li, Bo Wuhan University Key Laboratory of Aerospace Information Security and Trusted Computing Ministry of Education School of Cyber Science and Engineering Wuhan430072 China Wuhan University RiZhao Information Technology Institute Rizhao276800 China Huazhong University of Science and Technology National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Lab/Cluster and Grid Computing Lab School of Computer Science and Technology Wuhan430074 China Shanghai Jiao Tong University School of Cyber Science and Engineering Shanghai200240 China Hong Kong University of Science and Technology Department of Computer Science and Engineering Hong Kong

Cross-silo federated learning (FL) enables multiple institutions (clients) to collaboratively build a global model without sharing private data. To prevent privacy leakage during aggregation, homomorphic encryption (HE) is widely used to encrypt model updates, yet incurs high computation and communication overheads. To reduce these overheads, packed HE (PHE) has been proposed to encrypt multiple plaintexts into a single ciphertext. However, the original design of PHE assumes all clients share a single private key, making the system vulnerable to security threats of ciphertexts being intercepted and decrypted by honest-but-curious clients. Also, it does not consider the heterogeneity among different clients, resulting in undermined training efficiency with slow convergence and stragglers. To address these challenges, we propose FedPHE, a secure and efficient FL framework with PHE by jointly exploiting contribution-aware secure aggregation and straggler-resistant client selection. Using CKKS with sparsification and blinding, FedPHE achieves efficient secure aggregation that allows clients to only provide obscured encrypted updates while the server can perform aggregation by accounting for contributions of local updates. To mitigate the straggler effect, we devise a perturbed sketch-based selection to cherry-pick representative clients with heterogeneous models and computing capabilities in a communication-efficient and privacy-preserving manner. We show, through rigorous security analysis and extensive experiments, that FedPHE can efficiently safeguard clients' privacy, achieve 2.45-6.56× training speedup, cut the communication overhead by 1.32-24.85×, and reduce straggler effects by 1.89-2.78×. © 2004-2012 IEEE.

关键词： Privacy by design

来源：评论

学校读者我要写书评

暂无评论

Evaluating the point cloud of individual trees generated from images based on Neural Radiance fields (NeRF) method

arXiv

引用

arXiv 2023年

作者： Huang, Hongyu Tian, Guoji Chen, Chongcheng National Engineering Research Center of Geospatial Information Technology Fuzhou University Fuzhou350108 China Key Lab of Spatial Data Mining and Information Sharing of Ministry of Education Fuzhou University Fuzhou350108 China Fuzhou350108 China

Three-dimensional (3D) reconstruction of trees has always been a key task in precision forestry management and research. Due to the complex branch morphological structure of trees themselves and the occlusions from tree stems, branches and foliage, it is difficult to recreate a complete three-dimensional tree model from a two-dimensional image by conventional photogrammetric methods. In this study, based on tree images collected by various cameras in different ways, the Neural Radiance Fields (NeRF) method was used for individual tree reconstruction and the exported point cloud models are compared with point cloud derived from photogrammetric reconstruction and laser scanning methods. The results show that the NeRF method performs well in individual tree 3D reconstruction, as it has higher successful reconstruction rate, better reconstruction in the canopy area, it requires less amount of images as input. Compared with photogrammetric reconstruction method, NeRF has significant advantages in reconstruction efficiency and is adaptable to complex scenes, but the generated point cloud tends to be noisy and low resolution. The accuracy of tree structural parameters (tree height and diameter at breast height) extracted from the photogrammetric point cloud is still higher than those of derived from the NeRF point cloud. The results of this study illustrate the great potential of NeRF method for individual tree reconstruction, and it provides new ideas and research directions for 3D reconstruction and visualization of complex forest scenes. Copyright © 2023, The Authors. All rights reserved.

关键词： Photogrammetry

来源：评论

学校读者我要写书评

暂无评论

Admission Control with Latency Considerations for 5G Mobile Edge Computing

Admission Control with Latency Considerations for 5G Mobile ...

引用

IEEE International Symposium on World of Wireless Mobile and Multimedia Networks (WoWMoM)

作者： Ye Zhang Wuyungerile Li Winston K.G. Seah Inner Mongolia Key Lab. of Wireless Networking & Mobile Computing Engineering Research Center of Ecological Big Data Ministry of Education School of Computer Science Inner Mongolia University Hohhot China Wireless Networks Research Group School of Engineering and Computer Science Victoria University of Wellington Wellington New Zealand

The fifth generation (5G) mobile network is a new generation of broadband mobile communication technology with the potential to address the increasing demands of new user services and applications that have stringent low latency and high bandwidth requirements. Besides the enhanced mobile broadband (eMBB) which is an evolution of broadband services from previous generations, the ultra reliable low latency communication (URLLC) service comes with stringent delay requirements that are needed to support new applications like autonomous vehicles, augmented/virtual reality, etc. Mobile or multiple access edge computing (MEC) emerged to provide services and computing resources for users at the network edge to provide faster access speeds and lower end-to-end delays. To better meet user needs and maximize resource utilization, network resources need to be allocated and managed efficiently. Admission control for user requests is one of the methods used that can effectively prevent network congestion, thereby improving the overall performance of the system. In this paper, we propose a RED-based Admission Control with Latency Considerations (REDAL) algorithm for user admission control that aims to increase throughput, meet users’ delay requirements and reduce packet discard rate. By explicitly accounting for user traffic delay constraints and bandwidth requirements, we are able to meet the strict delay constraints of URLLC traffic while meeting the bandwidth requirements of eMBB traffic. We validate our approach in an MEC scenario to demonstrate high resource utilization and also keeping the request discard rate below 20%.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Sequence-Aware Online Container Scheduling with Reinforcement Learning in Parked Vehicle Edge Computing

引用

IEEE Transactions on Vehicular Technology 2025年

作者： Wu, Jianqiu Guo, Jianxiong Tang, Zhiqing Luo, Chuanwen Wang, Tian Jia, Weijia Beijing Normal-Hong Kong Baptist University Guangdong Key Lab of AI and Multi-Modal Data Processing Department of Computer Science Zhuhai519087 China Beijing Normal University Advanced Institute of Natural Sciences Zhuhai519087 China Beijing Normal-Hong Kong Baptist University Guangdong Key Lab of AI and Multi-Modal Data Processing Zhuhai519087 China Beijing Forestry University School of Information Science and Technology Beijing100083 China Engineering Research Center for Forestry-Oriented Intelligent Information Processing of National Forestry and Grassland Administration Beijing100083 China

Intelligent vehicles, often parked for long periods, are ideally suited to serve as computational nodes to expand the Mobile Edge Computing (MEC) infrastructure, with containerization significantly enhancing the system's load balancing, self-healing, resource isolation, and security. However, fluctuations in task demand and frequent container image downloads during peak hours create high loads on containerized nodes, as multiple mobile devices offload tasks simultaneously, leading to significant processing delays. Many existing studies make the simplified assumption of predefined patterns of task arrivals, which overlooks this issue and makes suboptimal decisions. In this paper, we consider a Parked Vehicles (PVs)-extended MEC scenario, where multiple devices request services on PVs functioning as edge servers, all controlled by a central base station. Task arrivals follow observed patterns based on long-term trends, such as peak and off-peak periods, resembling realistic arrival patterns rather than predefined ones. To optimize task offloading by identifying these patterns, we propose the Sequence-Aware Task Scheduling (SATS) algorithm, which is a policy gradient-based deep reinforcement learning approach that integrates Transformer and LSTM architectures to capture patterns in time-series task arrivals and relationships between nodes in a collaborative and containerized environment, thereby enhancing the efficiency of online task scheduling. The primary objective of SATS is to optimize the task offloading policy and minimize delay and energy consumption for all devices and PVs. Extensive numerical comparisons against baselines demonstrate the effectiveness and advantages of our algorithm. © 1967-2012 IEEE.

关键词： Mobile edge computing

来源：评论

学校读者我要写书评

暂无评论

Divert more attention to vision-language tracking 22

Divert more attention to vision-language tracking

引用

Proceedings of the 36th International Conference on Neural Information Processing Systems

作者： Mingzhe Guo Zhipeng Zhang Heng Fan Liping Jing Beijing Key Lab of Traffic Data Analysis and Mining Beijing Jiaotong University DiDi Chuxing Beijing China Department of Computer Science and Engineering University of North Texas

ISBN: (纸本)9781713871088

Relying on Transformer for complex visual feature learning, object tracking has witnessed the new standard for state-of-the-arts (SOTAs). However, this advancement accompanies by larger training data and longer training period, making tracking increasingly expensive. In this paper, we demonstrate that the Transformer-reliance is not necessary and the pure ConvNets are still competitive and even better yet more economical and friendly in achieving SOTA tracking. Our solution is to unleash the power of multimodal vision-language (VL) tracking, simply using ConvNets. The essence lies in learning novel unified-adaptive VL representations with our modality mixer (ModaMixer) and asymmetrical ConvNet search. We show that our unified-adaptive VL representation, learned purely with the ConvNets, is a simple yet strong alternative to Transformer visual features, by unbelievably improving a CNN-based Siamese tracker by 14.5% in SUC on challenging LaSOT (50.7%→65.2%), even outperforming several Transformer-based SOTA trackers. Besides empirical results, we theoretically analyze our approach to evidence its effectiveness. By revealing the potential of VL representation, we expect the community to divert more attention to VL tracking and hope to open more possibilities for future tracking beyond Transformer.

关键词：

来源：评论

学校读者我要写书评

暂无评论

A Real-Time Demonstration of Bidirectional Multi-Format/Rate-Adjustable Integrated Laser Communication/Ranging for Satellite Communication

A Real-Time Demonstration of Bidirectional Multi-Format/Rate...

引用

Conference on Lasers and Electro-Optics (CLEO)

作者： Yizhou Wang Yuanxiang Wang Zhen Luo Linsheng Zhong Yang Zou Xueyuan Ao Jiale Liu Feng Fan Xiaoliang Li Xiaoxiao Dai Qi Yang Deming Liu Wuhan National Lab for Optoelectronics (WNLO) & National Engineering Laboratory for Next Generation Internet Access System School of Optical and Electronic Information Huazhong University of Science and Technology Wuhan China 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 Wuhan China Beijing Research Institute of Telemetry Beijing China Jinyinhu Laboratory Wuhan China

A multi-format/rate-adjustable integrated laser communication/ranging system is proposed, achieving adjustable rates between 0.625-10Gbit/s. Real-time experiment shows that at 1e-4 BER the ranging error is only 0.29% ... 详细信息

ISBN: (纸本)9798350369311

关键词： Laser theory Codes Sensitivity Lasers Measurement uncertainty Measurement by laser beam Electro-optic effects Real-time systems Distance measurement Satellite communications

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：