AISHELL-2: Transforming Mandarin ASR Research Into Industrial Scale

TL;DR

AISHELL-2 provides a large-scale Mandarin speech corpus and an industrial-grade ASR recipe, facilitating research and practical applications in Mandarin speech recognition with state-of-the-art techniques.

Contribution

This work introduces a 1000-hour Mandarin speech corpus and an improved ASR pipeline with industrial features, advancing both research resources and practical system development.

Findings

Released 1000-hour Mandarin speech corpus for research and industry.

Developed an improved ASR recipe with state-of-the-art techniques.

Supported various channels and robust features for industrial applications.

Abstract

AISHELL-1 is by far the largest open-source speech corpus available for Mandarin speech recognition research. It was released with a baseline system containing solid training and testing pipelines for Mandarin ASR. In AISHELL-2, 1000 hours of clean read-speech data from iOS is published, which is free for academic usage. On top of AISHELL-2 corpus, an improved recipe is developed and released, containing key components for industrial applications, such as Chinese word segmentation, flexible vocabulary expension and phone set transformation etc. Pipelines support various state-of-the-art techniques, such as time-delayed neural networks and Lattic-Free MMI objective funciton. In addition, we also release dev and test data from other channels(Android and Mic). For research community, we hope that AISHELL-2 corpus can be a solid resource for topics like transfer learning and robust ASR. For industry, we hope AISHELL-2 recipe can be a helpful reference for building meaningful industrial systems and products.