DynamicLip Dataset
收藏资源简介:
DynamicLip数据集由中国科学院信息工程研究所创建,旨在支持基于唇部动态特征的连续生物认证研究。该数据集包含50名受试者的唇部动态信息,数据来源于受试者在不同环境下朗读特定短语时的唇部运动。数据集的设计充分考虑了自然语音中的唇部动态变化,涵盖了常见的音素和唇部运动模式。通过捕捉唇部的全局和局部特征,数据集为研究唇部动态认证提供了丰富的实验基础。该数据集的应用领域包括虚拟现实、智能车辆等高安全性和隐私要求的场景,旨在解决传统生物认证方法在动态环境中的局限性问题。
The DynamicLip dataset was created by the Institute of Information Engineering, Chinese Academy of Sciences, aiming to support research on continuous biometric authentication based on lip dynamic features. This dataset contains lip dynamic information from 50 subjects, with data collected from their lip movements when reading specific phrases in different environments. The design of the dataset fully considers the dynamic lip changes in natural speech, covering common phonemes and lip movement patterns. By capturing both global and local features of the lips, the dataset provides a rich experimental foundation for research on lip dynamic authentication. Its application scenarios include high-security and privacy-sensitive fields such as virtual reality and intelligent vehicles, aiming to address the limitations of traditional biometric authentication methods in dynamic environments.




