A Novel Labeled Human Voice Signal Dataset for Misbehavior Detection
收藏资源简介:
本数据集由拉合尔大学的软件工程系创建,专注于标记人类语音信号以检测不当行为。数据集通过实时收集参与者对心理学问题的回答,分为“正常”和“不当”两类。数据集内容包括12个问题的语音记录,通过高保真录音设备采集,并使用先进的信号处理技术进行分类。创建过程中,参与者需在两种不同的语音模式下回答问题,以捕捉不同的语音行为。该数据集主要应用于语音信号分析和机器学习领域,旨在提高语音识别技术的准确性和上下文感知能力。
This dataset was created by the Department of Software Engineering at the University of Lahore, focusing on annotating human speech signals for misconduct detection. It is collected in real time by gathering participants' responses to psychological questions, and categorized into two classes: "normal" and "inappropriate". The dataset includes speech recordings of 12 questions, acquired using high-fidelity audio recording equipment and classified via advanced signal processing technologies. During the dataset creation process, participants were required to answer questions in two distinct speech modes to capture diverse speech behaviors. This dataset is primarily applied in the fields of speech signal analysis and machine learning, aiming to improve the accuracy and context awareness of speech recognition technologies.




