thu-spmi/librispeech-phoneme-labels

---
license: apache-2.0
task_categories:
- automatic-speech-recognition
- translation
language:
- en
tags:
- lecture
pretty_name: d
size_categories:
- 10M<n<100M
---

# LibriSpeech IPA Phoneme Labels

This repository provides **IPA-based phoneme annotations and lexicon** for the LibriSpeech dataset.

All phoneme labels are converted from **CMU Pronouncing Dictionary (CMU-Dict)** phonemes into **IPA symbols** using deterministic rules, with the help of the following toolkit:

- https://pypi.org/project/pinyin-to-ipa

The data is intended for **phoneme-based ASR**, **P2G/G2P research**, **phoneme CTC / AED models**, and **cross-lingual phoneme experiments**.

## Dataset Structure

- `train-clean-100-phoneme`
- `train-clean-360-phoneme`
- `train-other-500-phoneme`
- `dev-clean-phoneme`
- `dev-other-phoneme`
- `test-clean-phoneme`
- `test-other-phoneme`
- `lexicon.txt`
- `phone_list`

许可证：Apache-2.0
任务类别：
- 自动语音识别（Automatic Speech Recognition, ASR）
- 机器翻译
语言：
- 英语
标签：
- 讲座
友好名称：d
数据规模分类：
- 10M<n<100M

# LibriSpeech 国际音标（International Phonetic Alphabet, IPA）音素标注集

本数据集仓库为LibriSpeech数据集提供**基于国际音标（IPA）的音素标注与发音词典**。

所有音素标注均通过确定性转换规则，借助下述工具包，将**CMU发音词典（CMU Pronouncing Dictionary, CMU-Dict）**的音素转换为国际音标符号：
- https://pypi.org/project/pinyin-to-ipa

本数据集可应用于面向音素的自动语音识别（ASR）、音素-字形转换（P2G）与字形-音素转换（G2P）研究、音素级连接时序分类（CTC）与注意力端到端（AED）模型研发，以及跨语言音素相关实验。

## 数据集结构

- `train-clean-100-phoneme`
- `train-clean-360-phoneme`
- `train-other-500-phoneme`
- `dev-clean-phoneme`
- `dev-other-phoneme`
- `test-clean-phoneme`
- `test-other-phoneme`
- `lexicon.txt`
- `phone_list`