Jove
Visualize
联系我们
JoVE
x logofacebook logolinkedin logoyoutube logo
关于 JoVE
概览领导团队博客JoVE 帮助中心
作者
出版流程编辑委员会范围与政策同行评审常见问题投稿
图书馆员
用户评价订阅访问资源图书馆顾问委员会常见问题
研究
JoVE JournalMethods CollectionsJoVE Encyclopedia of Experiments存档
教育
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab Manual教师资源中心教师网站
使用条款与条件
隐私政策
政策

相关概念视频

您也可能阅读

相关文章

通过共同作者、期刊和引用图与本文相关的文章。

排序
Same author

The effects of high-intensity interval training on executive function in children and adolescents: a systematic review and meta-analysis.

Frontiers in psychology·2026
Same author

Prevalence and incidence of pemphigus in urban Chinese: A nationwide population-based study.

JAAD international·2026
Same author

Hydrogen evolution and dynamics in hydrogel electrochemical cells for ischemia-reperfusion therapy.

Nature chemical engineering·2026
Same author

Load identification method for pepper harvesting drum based on dynamic chaotic characteristics of vibration-torque coupling.

PloS one·2026
Same author

Efficacy and Safety of CT-Verified Deep Needling at the Foramen Acupoint for Lumbar Disc Herniation: A Randomized Controlled Trial.

Pain research & management·2026
Same author

Quantifying the Economic Value of AI-Assisted Diabetic Retinopathy Screening: A Meta-Analysis Using the Incremental Net Benefit Framework.

Value in health : the journal of the International Society for Pharmacoeconomics and Outcomes Research·2026

相关实验视频

Updated: May 13, 2025

Author Spotlight: Investigating the Impact of Emotional Prosodies on Voice Recognition and Perception
05:48

Author Spotlight: Investigating the Impact of Emotional Prosodies on Voice Recognition and Perception

Published on: August 9, 2024

1.3K

用修改的激活残余网络进行目标样本挖掘,用于扬声器验证.

Ji Chaoqun1, Chen Wei1, Ye Peng1

  • 1Wenzhou Business College, Wenzhou, Zhejiang, The People's Republic of China.

PloS one
|April 16, 2025
PubMed
概括

SphereSpeaker是一种自适应目标功能,通过解决Softmax的局限性来改善扬声器验证. 它提高了网络稳定性和通用性,与其他深度学习方法相比,实现了最低的等错率.

科学领域:

  • 语音处理和机器学习
  • 生物识别安全系统 生物识别安全系统

背景情况:

  • 传统的Softmax扬声器验证方法存在培训-验证差异和不平衡样本问题.
  • 积极和消极样本之间的重叠相似性得分减少了扬声器验证系统中的歧视性.
  • 深度神经网络面临着像梯度消失,爆炸和模型退化等挑战.

研究的目的:

  • 引入一个自适应目标函数,SphereSpeaker,以克服传统Softmax在扬声器验证中的局限性.
  • 提高扬声器验证模型的稳定性和一般化能力.
  • 提高积极和消极演讲者样本之间的区分能力.

主要方法:

  • 开发了SphereSpeaker,这是一个根据Softmax.max改进的超参数的自适应性目标函数.
  • 整合了三个不同的角边缘,以改进网络更新.
  • 使用剩余网络PReLu (ResNet-P) 架构来解决深度神经网络训练问题.

主要成果:

  • SphereSpeaker 显示了对扬声器验证任务的更好的适用性.
  • 该ResNet-P架构增强了网络稳定性和通用性.
  • 实验结果显示,与现有的深度神经网络方法相比,同等错误率最低.

更多相关视频

Foreign Accent and Forensic Speaker Identification in Voice Lineups: The Influence of Acoustic Features Based on Prosody
09:09

Foreign Accent and Forensic Speaker Identification in Voice Lineups: The Influence of Acoustic Features Based on Prosody

Published on: September 27, 2024

379
Author Spotlight: Impact of Intergenic Interactions on Disease-Identifying Dark Biomarkers
03:37

Author Spotlight: Impact of Intergenic Interactions on Disease-Identifying Dark Biomarkers

Published on: March 1, 2024

607

相关实验视频

Last Updated: May 13, 2025

Author Spotlight: Investigating the Impact of Emotional Prosodies on Voice Recognition and Perception
05:48

Author Spotlight: Investigating the Impact of Emotional Prosodies on Voice Recognition and Perception

Published on: August 9, 2024

1.3K
Foreign Accent and Forensic Speaker Identification in Voice Lineups: The Influence of Acoustic Features Based on Prosody
09:09

Foreign Accent and Forensic Speaker Identification in Voice Lineups: The Influence of Acoustic Features Based on Prosody

Published on: September 27, 2024

379
Author Spotlight: Impact of Intergenic Interactions on Disease-Identifying Dark Biomarkers
03:37

Author Spotlight: Impact of Intergenic Interactions on Disease-Identifying Dark Biomarkers

Published on: March 1, 2024

607

结论:

  • SphereSpeaker有效地解决了培训验证差异和演讲者验证中的样本不平衡问题.
  • 拟议的ResNet-P与SphereSpeaker显著提高了扬声器验证系统的性能.
  • 这种方法为语音识别和验证提供了更强大,更准确的解决方案.