Open Source MagicData-RAMC: A Rich Annotated Mandarin Conversational(RAMC) Speech Dataset

TL;DR

This paper presents MagicData-RAMC, a comprehensive high-quality Mandarin conversational speech dataset with extensive annotations, designed to facilitate research in various speech processing tasks such as recognition, diarization, and topic detection.

Contribution

The paper introduces MagicData-RAMC, a new large-scale, richly annotated Mandarin conversational speech dataset with diverse topics and detailed speaker information, filling a gap in Mandarin speech resources.

Findings

Dataset contains 180 hours of speech data from native speakers.

Includes detailed annotations like speaker activity, topics, and speaker info.

Provides baseline experimental results for multiple speech tasks.

Abstract

This paper introduces a high-quality rich annotated Mandarin conversational (RAMC) speech dataset called MagicData-RAMC. The MagicData-RAMC corpus contains 180 hours of conversational speech data recorded from native speakers of Mandarin Chinese over mobile phones with a sampling rate of 16 kHz. The dialogs in MagicData-RAMC are classified into 15 diversified domains and tagged with topic labels, ranging from science and technology to ordinary life. Accurate transcription and precise speaker voice activity timestamps are manually labeled for each sample. Speakers' detailed information is also provided. As a Mandarin speech dataset designed for dialog scenarios with high quality and rich annotations, MagicData-RAMC enriches the data diversity in the Mandarin speech community and allows extensive research on a series of speech-related tasks, including automatic speech recognition, speaker diarization, topic detection, keyword search, text-to-speech, etc. We also conduct several relevant tasks and provide experimental results to help evaluate the dataset.