Multilingual TEDx Corpus
收藏资源简介:
Multilingual TEDx Corpus是由约翰斯霍普金斯大学创建的一个多语言数据集,旨在支持多种非英语源语言的语音识别(ASR)和语音翻译(ST)研究。该数据集包含来自8种源语言的TEDx演讲音频记录,并提供了6种目标语言的翻译。数据集的创建过程包括将转录文本分割成句子,并与源语言音频和目标语言翻译进行对齐。该数据集的应用领域广泛,旨在解决多语言环境下的语音识别和翻译问题,特别关注低资源语言对的翻译性能提升。
Multilingual TEDx Corpus is a multilingual dataset developed by Johns Hopkins University, designed to support research on automatic speech recognition (ASR) and speech translation (ST) across multiple non-English source languages. This dataset includes audio recordings of TEDx talks from 8 source languages, with translations provided for 6 target languages. The dataset construction process involves segmenting the transcribed text into sentences and aligning them with the source-language audio and target-language translations. The dataset has a wide range of applications, aiming to address speech recognition and translation challenges in multilingual environments, with a particular focus on improving translation performance for low-resource language pairs.

- 1The Multilingual TEDx Corpus for Speech Recognition and Translation约翰斯霍普金斯大学 · 2021年



