<b>A Pilot Speech Corpus for Studying Device and Environmental Variability in Voice Biometrics</b>

This dataset provides a curated pilot corpus for studying <b>device and environmental variability in voice biometrics</b>. It contains <b>480 speech recordings</b> from <b>12 participants</b> (Japan, Nigeria, Ivory Coast, France, Germany, and Indonesia), each contributing 40 utterances recorded across multiple devices and environments.Recordings were made using the Samsung A04s, OnePlus Nord (both direct and in-call), iPhone 15 Pro, and a USB condenser microphone (connected to a <b>MacBook)</b>, under both <b>indoor (semi-controlled lobby)</b> and <b>outdoor (campus)</b> conditions. All files are stored in <b>WAV format (8–16 kHz, 16-bit PCM)</b>, accompanied by a <b>metadata file (CSV/Excel)</b> with anonymized attributes such as nationality, gender, age, and English proficiency.The dataset supports research in <b>speech enhancement</b> (spectral subtraction, Wiener filtering, adaptive filtering), <b>speaker identification and verification</b>, <b>spoofing resilience</b>, and <b>liveness detection</b>. Validation experiments confirmed that adaptive filtering achieved the highest accuracy (97%), highlighting both the challenges of cross-device variability and the potential for robust enhancement methods.This corpus provides a valuable benchmark for developing <b>secure and consistent voice biometric systems</b>, particularly in real-world applications such as <b>mobile banking authentication</b> and <b>low-resource environments</b>.本数据集为研究**语音生物识别中的设备与环境变异性（voice biometrics）**提供了经过精选的试点语料库。该数据集包含来自12名参与者的**480条语音录音（speech recordings）**，参与者涵盖日本、尼日利亚、科特迪瓦、法国、德国与印度尼西亚，每位参与者录制了40条语音话语片段，采集场景覆盖多种设备与环境。录音所用设备包括三星A04s、OnePlus Nord（支持直录与通话中录制两种模式）、iPhone 15 Pro，以及连接至**MacBook**的USB电容麦克风；采集环境分为**室内（半受控大堂）**与**室外（校园）**两类场景。所有音频文件均以**WAV格式（8–16 kHz，16位PCM）**存储，同时附带包含匿名化属性的**元数据文件（CSV/Excel）**，属性涵盖国籍、性别、年龄与英语水平。本数据集可支撑以下方向的研究：**语音增强（speech enhancement）**（包括谱减法、维纳滤波、自适应滤波）、**说话人识别与验证**、**抗欺骗性（spoofing resilience）**以及**活体检测（liveness detection）**。验证实验结果显示，自适应滤波的识别准确率最高（97%），这既凸显了跨设备变异性带来的研究挑战，也验证了鲁棒性增强方法的应用潜力。该语料库可为开发**安全且一致的语音生物识别系统**提供极具价值的基准测试集，尤其适用于移动银行身份验证、低资源环境等真实应用场景。