Which phonemes convey more speaker traits is a long-standing question, and various perception experiments were conducted with human subjects. For speaker recognition, studies were conducted with the conventional statistical models and the drawn conclusions are more or less consistent with the perception results. However, which phonemes are more important with modern deep neural models is still unexplored, due to the opaqueness of the decision process. This paper conducts a novel study for the attribution of phonemes with two types of deep speaker models that are based on TDNN and CNN respectively, from the perspective of model explanation. Specifically, we conducted the study by two post-explanation methods: LayerCAM and Time Align Occlusion (TAO). Experimental results showed that: (1) At the population level, vowels are more important than consonants, confirming the human perception studies. However, fricatives are among the most unimportant phonemes, which contrasts with previous studies. (2) At the speaker level, a large between-speaker variation is observed regarding phoneme importance, indicating that whether a phoneme is important or not is largely speaker-dependent.
Phoneme-aware and Channel-wise Attentive Learning for Text DependentSpeaker Verification
Xiamen University（厦门大学）
本文提出了一个多任务学习网络与音素意识和通道的注意学习策略的文本相关的说话人确认（SV）。该结构采用帧级多任务学习和分段级对抗学习相结合的方法进行说话人嵌入提取。在主网络中的帧级特征上利用音素感知注意池用于说话人分类器，在辅助子网中具有对应的音素分布的后验概率。此外，挤压和激励（SE块）的引入执行动态通道特征重新校准，这提高了表示能力。该方法利用了说话人与密码短语的特性，并分别从时间和通道两个方面通过音素感知集中和SE块来进一步改进。在RSR2015 Part 1数据库上进行的实验证实，所提出的系统在文本相关SV上取得了出色的效果。
Analysis of Speech Temporal Dynamics in the Context of Speaker Verification and Voice Anonymization
Université de Lorraine, CNRS, Inria Loria（卢瓦尔大学、法国国家科学研究中心、Inria Loria）；Université de Lille, CNRS, Inria Centrale Lille, UMR 9189 - CRIStAL Lille（里尔大学、法国国家科学研究中心、Inria Centrale Lille、UMR 9189 - CRIStAL Lille）
ProLombard: Structured Multi-Scale Modeling for Normal-to-Lombard Speech Conversion
National Engineering Research Center for Multimedia Software（国家多媒体软件工程技术研究中心）；School of Computer Science, Wuhan University（武汉大学计算机学院）；Hubei Key Laboratory of Multimedia and Network Communication Engineering（多媒体网络通信工程湖北省重点实验室）；Guangdong OPPO Mobile Telecommunications Corp.（广东欧珀移动通信有限公司）
Stabilizing Short Duration Speaker Verification through Neural Re-scoring with Hybrid Enrollment
Shanghai University, China（上海大学）；Xi’an Jiaotong-Liverpool University, China（西安交通大学利物浦大学）；Hithink RoyalFlush AI Research Institute, China（Hithink皇家Flush人工智能研究院）
MGFF-TDNN: A Multi-Granularity Feature Fusion TDNN Model with Depth-Wise Separable Module for Speaker Verification
College of Computer Science（计算机科学学院）；South-Central Minzu University（西南民族大学）；Wuhan Dongxin Tongbang Information Technology Co., Ltd.（武汉东信通 bang信息技术有限公司）
Speech2Phone: A Novel and Efficient Method for Training Speaker Recognition Models
University of São Paulo（圣保罗大学）；Federal University of Technology - Paraná（巴拉那联邦理工大学）；DefinedCrowd Corp.（DefinedCrowd公司）；Federal University of Goias（戈亚斯联邦大学）
Learning Explicit Prosody Models and Deep Speaker Embeddings for Atypical Voice Conversion
Human-Computer Communications Laboratory（人机通信实验室）；The Chinese University of Hong Kong（香港中文大学）；SpeechX Limited（SpeechX有限公司）；University of Cambridge（剑桥大学）
Deep neural network based i-vector mapping for speaker verification using short utterances
University of California, Los Angeles（加利福尼亚大学洛杉矶分校）；Snap Inc（Snap公司）
由于短话语之间的大变化和内容失配，使用短话语的文本无关的说话人识别是一个非常具有挑战性的任务。基于I-向量的系统已经成为说话人确认应用的标准，但是它们对于短话语不太有效。在本文中，我们首先比较了两种最先进的通用背景模型训练方法，用于使用全长和短话语评估任务进行i向量建模。这两种方法是基于高斯混合模型（GMM）和基于深度神经网络（DNN）的方法。结果表明，I-vector_DNN系统优于I-vector_GMM系统在不同的持续时间。然而，这两个系统的性能显着降低的话语的持续时间减少。为了解决这个问题，我们提出了两种新的非线性映射方法，训练DNN模型将从短话语中提取的i向量映射到其相应的长话语i向量。映射后的i向量能够恢复缺失信息，并减小原短话语i向量的方差。所提出的方法都模型的联合表示的短和长的话语i-向量使用自动编码器。使用NIST SRE 2010数据集的实验结果表明，当使用具有残差块的深度编码器并添加额外的音素向量时，这两种方法都提供了显着的改进，并导致与基线系统相比相等错误率的最大28.43%的相对改进。当在Speaker In The Wild数据集上进一步测试SRE 10的最佳验证模型时，该方法在任意持续时间（1-5秒）的短话语条件下提高了23.12%。
Probing phoneme, language and speaker information in unsupervised speech representations
Cognitive Machine Learning (ENS–CNRS–EHESS–INRIA–PSL Research University)（认知机器学习实验室（巴黎高等师范学院-法国国家科学研究中心-法国高等社会科学研究院-法国国家信息与自动化研究所-巴黎文理研究大学））；Université de Paris Cité（巴黎西岱大学）；CNRS（法国国家科学研究中心）；Meta AI Research（Meta人工智能研究院）
Two-Stream Joint-Training for Speaker Independent Acoustic-to-Articulatory Inversion
Tianjin University（天津大学）；Tianjin International Engineering Institute, Tianjin University（天津大学天津国际工程学院）；The Hong Kong University of Science and Technology (Guangzhou)（香港科技大学（广州））；Institute of Linguistics, Chinese Academy of Social Sciences（中国社会科学院语言研究所）
Profiling the Voice: Speaker-Specific Phoneme Fingerprinting for Speech Deepfake Detection
Key Laboratory of Aerospace Information Security and Trusted Computing, Ministry of Education（航空航天信息安全与可信计算重点实验室，教育部）；School of Cyber Science and Engineering, Wuhan University（武汉大学计算机科学与工程学院）
FabasedVC: Enhancing Voice Conversion with Text Modality Fusion and Phoneme-Level SSL Features
School of Software Engineering, Xi'an Jiaotong University（西安交通大学软件工程学院）；AI Platform Department, bilibili（bilibili人工智能平台部）；School of Software Engineering, East China Normal University（华东师范大学软件工程学院）；School of Computer Science and Artificial Intelligence, Fudan University（复旦大学计算机科学与人工智能学院）；Division of Music and Audio, Union Wheatland Culture and Media Ltd.（Union Wheatland Culture and Media Ltd.音乐与音频部）
Decoding Speaker-Normalized Pitch from EEG for Mandarin Perception
University of Science and Technology of China（中国科学技术大学）；National Engineering Research Center of Speech and Language Information Processing（语音与语言信息处理国家级工程研究中心）；Interdisciplinary Research Center for Linguistic Sciences（语言科学交叉研究中心）
Enhancing Quranic Learning: A Multimodal Deep Learning Approach for Arabic Phoneme Recognition
Department of Electronics and Communication Engineering, Kocaeli University, 41001, Kocaeli, Turkey（电子与通信工程系，科拉尔大学）；Department of Information Systems Engineering, Kocaeli University, 41001, Kocaeli, Turkey（信息系统工程系，科拉尔大学）；Maviay Consultancy Company, Kocaeli University Technopark, 41275, Kocaeli, Turkey（Maviay咨询公司，科拉尔大学技术园区）
The FruitShell French synthesis system at the Blizzard 2023 Challenge
State Key Laboratory of Multimodal Artificial Intelligence Systems, CASIA（中国科学院自动化研究所多模态人工智能系统国家重点实验室）；School of Artificial Intelligence, University of Chinese Academy of Sciences（中国科学院大学人工智能学院）；Anhui University（安徽大学）；Shanghai Polytechnic University（上海第二工业大学）
