检索结果-内蒙古大学图书馆

IEEE Spoken Language Technology Workshop

作者： Hankun Wang Chenpeng Du Yiwei Guo Shuai Wang Xie Chen Kai Yu Department of Computer Science and Engineering MoE Key Lab of Artificial Intelligence AI Institute X-LANCE Lab Shanghai Jiao Tong University Shanghai China Shenzhen Research Institute of Big Data The Chinese University of Hong Kong Shenzhen China

ISBN: (数字)9798350392258

ISBN: (纸本)9798350392265

Recent popular decoder-only text-to-speech models are known for their ability of generating natural-sounding speech. However, such models sometimes suffer from word skipping and repeating due to the lack of explicit monotonic alignment constraints. In this paper, we notice from the attention maps that some particular attention heads of the decoder-only model indicate the alignments between speech and text. We call the attention maps of those heads Alignment-Emerged Attention Maps (AEAMs). Based on this discovery, we propose a novel inference method without altering the training process, named Attention-Constrained Inference (ACI), to facilitate monotonic synthesis. It first identifies AEAMs using the Attention Sweeping algorithm and then applies constraining masks on AEAMs. Our experimental results on decoderonly TTS model VALL-E show that the WER of synthesized speech is reduced by up to 20.5% relatively with ACI while the naturalness and speaker similarity are comparable.

关键词： Training Conferences Robustness Inference algorithms Text to speech Optimization

来源：评论

学校读者我要写书评

暂无评论

Prototype and Instance Contrastive Learning for Unsupervised Domain Adaptation in Speaker Verification

Prototype and Instance Contrastive Learning for Unsupervised...

引用

International Symposium on Chinese Spoken Language Processing

作者： Wen Huang Bing Han Zhengyang Chen Shuai Wang 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 Shenzhen Research Institute of Big Data The Chinese University of Hong Kong Shenzhen

ISBN: (数字)9798331516826

ISBN: (纸本)9798331516833

Speaker verification system trained on one domain usually suffers performance degradation when applied to another domain. To address this challenge, researchers commonly use feature distribution matching-based methods in unsupervised domain adaptation scenarios where some unlabeled target domain data is available. However, these methods often have limited performance improvement and lack generalization in various mismatch situations. In this paper, we propose Prototype and Instance Contrastive Learning (PICL), a novel method for unsupervised domain adaptation in speaker verification through dual-level contrastive learning. For prototype contrastive learning, we generate pseudo labels via clustering to create dynamically updated prototype representations, aligning instances with their corresponding class or cluster prototypes. For instance contrastive learning, we minimize the distance between different views or augmentations of the same instance, ensuring robust and invariant representations resilient to variations like noise. This dual-level approach provides both high-level and low-level supervision, leading to improved generalization and robustness of the speaker verification model. Unlike previous studies that only evaluated mismatches in one situation, we have conducted relevant explorations on various datasets and achieved state-of-the-art performance currently, which also proves the generalization of our method.

关键词： Degradation Adaptation models Noise Prototypes Contrastive learning Robustness

来源：评论

学校读者我要写书评

暂无评论

Totally asymmetric simple exclusion process on a dynamic lattice with local inhomogeneity

引用

Results in Physics 2024年 64卷

作者： Lin, Congping Huang, Yuesong Xu, Zihui Zhang, Yiwei School of Mathematics and Statistics Huazhong University of Science and Technology China Hubei Key Lab of Engineering Modelling and Scientific Huazhong University of Science and Technology China School of Mathematics and Big Data Anhui University of Science and Technology China

In this manuscript, we introduce a totally asymmetric simple exclusion process on a dynamic lattice with inhomogeneity to model motor-based long-range transport along microtubules (MTs) which exhibit cycles of growth, shortening and regrowth at their growing tips. By mean-field approximation and numerical simulations, we explore phase diagrams for the particle density near the dynamic end of the lattice as well as over the entire lattice. In particular, we find seven different phases for the density over the entire lattice and explore the corresponding phase diagram by analytical analysis. Interestingly, we find that the density may have a linear profile in part of the lattice. When the lattice has a defect, we separate the lattice into two subsystems: a right subsystem with fixed length and a defect in the middle and a left subsystem with dynamic length. When the right subsystem is in maximum-current phase, we calculate an analytical approximation of the density away from the boundaries and the defect site and the density is in association with the reduced hopping rate at the defect site;the density at the left side of the defect site serves as the effective entry rate of the left subsystem. We find that with defects, the linear profiles no longer appear at the left end of the left subsystem and only five possible phases could occur. This is a step-forward to a more comprehensive mathematical description of intracellular long-range transport of organelles along MTs with instability. © 2024

关键词： Dynamic lattice Inhomogeneity Phase diagram TASEP

来源：评论

学校读者我要写书评

暂无评论

Validating SMT Solvers via Skeleton Enumeration Empowered by Historical Bug-Triggering Inputs

Validating SMT Solvers via Skeleton Enumeration Empowered by...

引用

International Conference on Software Engineering (ICSE)

作者： Maolin Sun Yibiao Yang Ming Wen Yongcong Wang Yuming Zhou Hai Jin School of Cyber Science and Engineering Huazhong University of Science and Technology China Hubei Key Laboratory of Distributed System Security Services Computing Technology and System Lab Cluster and Grid Computing Lab Hubei Engineering Research Center on Big Data Security National Engineering Research Center for Big Data Technology and System State Key Laboratory for Novel Software Technology Nanjing University China School of Computer Science and Technology Huazhong University of Science and Technology China

SMT solvers check the satisfiability of logic formulas over first-order theories, which have been utilized in a rich number of critical applications, such as software verification, test case generation, and program synthesis. Bugs hidden in SMT solvers would severely mislead those applications and further cause severe consequences. Therefore, ensuring the reliability and robustness of SMT solvers is of critical importance. Although many approaches have been proposed to test SMT solvers, it is still a challenge to discover bugs effectively. To tackle such a challenge, we conduct an empirical study on the historical bug-triggering formulas in SMT solvers' bug tracking systems. We observe that the historical bug-triggering formulas contain valuable skeletons (i.e., core structures of formulas) as well as associated atomic formulas which can cast significant impacts on formulas' ability in triggering bugs. Therefore, we propose a novel approach that utilizes the skeletons extracted from the historical bug-triggering formulas and enumerates atomic formulas under the guidance of association rules derived from historical formulas. In this study, we realized our approach as a practical fuzzing tool HistFuzz and conducted extensive testing on the well-known SMT solvers Z3 and cvc5. To date, HistFuzz has found 111 confirmed new bugs for Z3 and cvc5, of which 108 have been fixed by the developers. More notably, out of the confirmed bugs, 23 are soundness bugs and invalid model bugs found in the solvers' default mode, which are essential for SMT solvers. In addition, our experiments also demonstrate that HistFuzz outperforms the state-of-the-art SMT solver fuzzers in terms of achieved code coverage and effectiveness.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Detecting Backdoors During the Inference Stage Based on Corruption Robustness Consistency

Detecting Backdoors During the Inference Stage Based on Corr...

引用

Conference on Computer Vision and Pattern Recognition (CVPR)

作者： Xiaogeng Liu Minghui Li Haoyu Wang Shengshan Hu Dengpan Ye Hai Jin Libing Wu Chaowei Xiao School of Cyber Science and Engineering Huazhong University of Science and Technology National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Lab Hubei Key Laboratory of Distributed System Security Hubei Engineering Research Center on Big Data Security School of Software Engineering Huazhong University of Science and Technology School of Cyber Science and Engineering Wuhan University School of Computer Science and Technology Huazhong University of Science and Technology Cluster and Grid Computing Lab Arizona State University

Deep neural networks are proven to be vulnerable to backdoor attacks. Detecting the trigger samples during the inference stage, i.e., the test-time trigger sample detection, can prevent the backdoor from being triggered. However, existing detection methods often require the defenders to have high accessibility to victim models, extra clean data, or knowledge about the appearance of backdoor triggers, limiting their practicality. In this paper, we propose the test-time corruption robustness consistency evaluation (TeCo) 1 1 https://***/CGCL-codes/TeCo, a novel test-time trigger sample detection method that only needs the hard-label outputs of the victim models without any extra information. Our journey begins with the intriguing observation that the backdoor-infected models have similar performance across different image corruptions for the clean images, but perform discrepantly for the trigger samples. Based on this phenomenon, we design TeCo to evaluate test-time robustness consistency by calculating the deviation of severity that leads to predictions' transition across different corruptions. Extensive experiments demonstrate that compared with state-of-the-art defenses, which even require either certain information about the trigger types or accessibility of clean data, TeCo outperforms them on different backdoor attacks, datasets, and model architectures, enjoying a higher AUROC by 10% and 5 times of stability.

关键词：

来源：评论

学校读者我要写书评

暂无评论

UniSleepPos: Sleep Posture Identification System Utilizing Millimeter-wave Radar

UniSleepPos: Sleep Posture Identification System Utilizing M...

引用

IEEE International Conference on Bioinformatics and Biomedicine (BIBM)

作者： Kun Liu Min Li Chu He Xu Tian Junbin Mao Min Zeng Jin Liu Xinjiang Engineering Research Center of Big Data and Intelligent Software School of Software Xinjiang University Urumqi China Hunan Provincial Key Lab on Bioinformatics School of Computer Science and Engineering Central South University Changsha China

ISBN: (数字)9798350386226

ISBN: (纸本)9798350386233

Sleep posture identification is crucial for accurately assessing sleep quality and diagnosing related diseases. In the realm of non-intrusive sleep monitoring, non-contact technologies are becoming increasingly mainstream. Millimeter-wave radar is frequently utilized in sleep posture identification due to its high resolution, strong penetration, and excellent sensitivity. However, traditional radar-based methods for sleep posture identification often struggle with reliability when dealing with diverse individuals and complex sleep environments. To address these challenges, we propose UniSleepPos, which designs a novel dual-view fusion mechanism to integrate depression and elevation angle signals obtained from radar, thus accurately capturing the posture information of the monitored subject in three-dimensional space. Furthermore, we combine sleep posture identification with individual characteristics, utilizing existing individual labels as prior knowledge to assist in sleep posture identification. The integration of prior knowledge provides a valuable information source for the model, helping to enhance its understanding of the data and improve its performance. We collected sleep posture data from eight volunteers using millimeter-wave radar devices under various environmental conditions. Leave-one-subject-out experiments were conducted to validate the effectiveness of UniSleepPos. The results indicated that UniSleepPos significantly outperforms existing methods, demonstrating its potential for practical applications.

关键词： Performance evaluation Sensitivity Spaceborne radar Millimeter wave radar Depression data models Reliability Monitoring Signal resolution Diseases

来源：评论

学校读者我要写书评

暂无评论

No Free Lunch Theorem for Privacy-Preserving LLM Inference

arXiv

引用

arXiv 2024年

作者： Zhang, Xiaojin Fei, Yulin Kang, Yan Chen, Wei Fan, Lixin Jin, Hai Yang, Qiang Huazhong University of Science and Technology China WeBank China Hong Kong University of Science and Technology China 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 Huazhong University of Science and Technology Wuhan430074 China

Individuals and businesses have been significantly benefited by Large Language Models (LLMs) including PaLM, Gemini and ChatGPT in various ways. For example, LLMs enhance productivity, reduce costs, and enable us to focus on more valuable tasks. Furthermore, LLMs possess the capacity to sift through extensive datasets, uncover underlying patterns, and furnish critical insights that propel the frontiers of technology and science. However, LLMs also pose privacy concerns. Users’ interactions with LLMs may expose their sensitive personal or company information. A lack of robust privacy safeguards and legal frameworks could permit the unwarranted intrusion or improper handling of individual data, thereby risking infringements of privacy and the theft of personal identities. To ensure privacy, it is essential to minimize the dependency between shared prompts and private information. Various randomization approaches have been proposed to protect prompts’ privacy, but they may incur utility loss compared to unprotected LLMs prompting. Therefore, it is essential to evaluate the balance between the risk of privacy leakage and loss of utility when conducting effective protection mechanisms. The current study develops a framework for inferring privacy-protected Large Language Models (LLMs) and lays down a solid theoretical basis for examining the interplay between privacy preservation and utility. The core insight is encapsulated within a theorem that is called as the NFL (abbreviation of the word No-Free-Lunch) Theorem. Copyright © 2024, The Authors. All rights reserved.

关键词： Computational linguistics

来源：评论

学校读者我要写书评

暂无评论

Challenges and Approaches for Mitigating Byzantine Attacks in Federated Learning

Challenges and Approaches for Mitigating Byzantine Attacks i...

引用

IEEE International Conference on Trust, Security and Privacy in Computing and Communications (TrustCom)

作者： Junyu Shi Wei Wan Shengshan Hu Jianrong Lu Leo Yu Zhang School of Cyber Science and Engineering Huazhong University of Science and Technology National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Lab Hubei Engineering Research Center on Big Data Security School of Information Technology Deakin University

Recently emerged federated learning (FL) is an attractive distributed learning framework in which numerous wireless end-user devices can train a global model with the data remained autochthonous. Compared with the traditional machine learning framework that collects user data for centralized storage, which brings huge communication burden and concerns about data privacy, this approach can not only save the network bandwidth but also protect the data privacy. Despite the promising prospect, Byzantine attack, an intractable threat in conventional distributed network, is discovered to be rather efficacious against FL as well. In this paper, we conduct a comprehensive investigation of the state-of-the-art strategies for defending against Byzantine attacks in FL. We first provide a taxonomy for the existing defense solutions according to the techniques they used, followed by an across-the-board comparison and discussion. Then we propose a new Byzantine attack method called weight attack to defeat those defense schemes, and conduct experiments to demonstrate its threat. The results show that existing defense solutions, although abundant, are still far from fully protecting FL. Finally, we indicate possible countermeasures for weight attack, and highlight several challenges and future research directions for mitigating Byzantine attacks in FL.

关键词： Wireless communication Training Privacy data privacy Federated learning Distance learning Taxonomy

来源：评论

学校读者我要写书评

暂无评论

The Entry-Extensible Cuckoo Filter 17th

The Entry-Extensible Cuckoo Filter

引用

17th IFIP WG 10.3 International Conference on Network and Parallel Computing, NPC 2020

作者： Yu, Shuiying Wu, Sijie Chen, Hanhua Jin, Hai 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 Huazhong University of Science and Technology Wuhan China

ISBN: (纸本)9783030794774

The emergence of large-scale dynamic sets in real applications brings severe challenges in approximate set representation structures. A dynamic set with changing cardinality requires an elastic capacity of the approximate set representation structure, while traditional static structures, e.g., bloom filter, cuckoo filter, and their variants cannot satisfy the requirements. Existing dynamic approximate set representation structures only provide filter-level extensions, which require a single membership query to probe all discrete filters one by one. The large number of small discrete memory accesses takes up the vast majority of query time and results in unsatisfied query performance. To address the problem, in this work we propose the entry-extensible cuckoo filter (E2CF) to reduce memory access overhead for dynamic set representation and accelerate the membership query. E2CF utilizes adjacent buckets with continuous physical addresses in a cuckoo filter to extend bucket entries, which avoids many discrete memory accesses in a query. To further make E2CF space and time efficient, we adopt asynchronous extension and fine-grained splitting methods. Experiment results show that compared to state-of-the-art designs, E2CF reduces the query and insertion time by 82% and 28%, respectively. © 2021, IFIP International Federation for Information Processing.

关键词： data structures

来源：评论

学校读者我要写书评

暂无评论

Aligning Language Models Using Follow-up Likelihood as Reward Signal

arXiv

引用

arXiv 2024年

作者： Zhang, Chen Chong, Dading Jiang, Feng Tang, Chengguang Gao, Anningzhe Tang, Guohua Li, Haizhou National University of Singapore Singapore Peking University China The Chinese University of Hong Kong Shenzhen China Tencent AI Lab China Shenzhen Research Institute of Big Data China University of Science and Technology of China China

In natural human-to-human conversations, participants often receive feedback signals from one another based on their follow-up reactions. These reactions can include verbal responses, facial expressions, changes in emotional state, and other non-verbal cues. Similarly, in human-machine interactions, the machine can leverage the user’s follow-up utterances as feedback signals to assess whether it has appropriately addressed the user’s request. Therefore, we propose using the likelihood of follow-up utterances as rewards to differentiate preferred responses from less favored ones, without relying on human or commercial LLM-based preference annotations. Our proposed reward mechanism, "Follow-up Likelihood as Reward" (FLR), matches the performance of strong reward models trained on large-scale human or GPT-4 annotated data on 8 pairwise-preference and 4 rating-based benchmarks. Building upon the FLR mechanism, we propose to automatically mine preference data from the online generations of a base policy model. The preference data are subsequently used to boost the helpfulness of the base model through direct alignment from preference (DAP) methods, such as direct preference optimization (DPO). Lastly, we demonstrate that fine-tuning the language model that provides follow-up likelihood with natural language feedback significantly enhances FLR’s performance on reward modeling benchmarks and effectiveness in aligning the base policy model’s helpfulness1 Copyright © 2024, The Authors. All rights reserved.

关键词： Benchmarking

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：