Microsoft Speech Language Translation (MSLT) Corpus (v1.1)

The Microsoft Speech Language Translation Corpus release contains conversational, bilingual speech test and tuning data for English, Chinese, and Japanese collected by Microsoft Research. The package includes audio data, transcripts, and translations and allows end-to-end testing of spoken language translation systems on real-world data. All data contained in this release has been created using a non-public version of Skype Translator. NO PRIVATE USER DATA HAS BEEN COLLECTED OR RELEASED. Instead we hired consultants to have loosely constrained conversations, giving them a list of predefined topics to talk about and a few related questions to start the conversations. Topical constraints were loosely enforced so as to ensure free-form conversations. See the MT Summit paper for more details. We release two sets, one containing Test data, the second containing Dev data. Each set contains data for three languages: English, Chinese, and Japanese. For every utterance, we include the audio file in WAVE format, the disfluent transcript, a cleaned up, segmented and fluent version of the transcript, and the translation from English into Chinese or Japanese or vice versa.

Supported Operating Systems

Android, Apple Mac OS X, Windows 10, Windows 8

• Android, Apple Mac OS X, Windows 8, Windows 10

• Click Download and follow the instructions.