147小时菲律宾语自然对话电话信道语音数据

他加禄语自然对话电话信道语音数据，基于约30个常见主题来模拟录制。此数据集标注了文本内容、句时间戳、说话人身份、性别等多种属性，由约264名来自不同地域和文化背景的菲律宾本土人录制，准确性高，为语音识别相关研究及应用提供了丰富的资源，经多家AI公司验证：有助于模型面对真实世界的多样性时能够表现出色。我们严格遵循数据保护法规和隐私规定，确保数据采集、存储和使用的过程中维护用户的隐私和合法权益，所有数据均遵循GDPR, CCPA, PIPL。

This is a Tagalog natural conversational telephone-channel speech dataset, recorded by simulating conversations centered on approximately 30 common topics. This dataset is annotated with multiple attributes including text transcripts, sentence-level timestamps, speaker identities, genders, and more. It was collected from approximately 264 native Filipino speakers with diverse regional and cultural backgrounds, boasting high data accuracy and providing abundant resources for speech recognition-related research and applications. Verified by multiple AI enterprises, this dataset helps models deliver outstanding performance when coping with the diversity of real-world scenarios. We strictly comply with data protection regulations and privacy provisions, ensuring that user privacy and legitimate rights and interests are safeguarded throughout the entire process of data collection, storage, and utilization. All datasets adhere to GDPR, CCPA, and PIPL.

搜集汇总

背景与挑战

背景概述

该数据集包含147小时的他加禄语自然对话电话信道语音数据，由264名菲律宾本土人录制，覆盖30个常见主题，标注了多种属性，适用于语音识别研究。数据格式为8kHz，8bit，ulaw/alaw pcm，单声道，词准确率达98%。