Introduction

The IWSLT Evaluation will focus this year on two tasks:

  • the translation of talks consisting of TED talks and talks from the QED Corpus.
  • the translation of conversations conducted via Skype, the Microsoft Speech Language (MSLT) test task

As in the previous years, the evaluation offers specific tracks for all the core technologies involved in spoken language translation, namely:

  • Automatic speech recognition (ASR), i.e. the conversion of a speech signal into a transcript,

  • Machine translation (MT), i.e. the translation of a polished transcript into another language,

  • Spoken language translation (SLT), i.e. the conversion and translation of a speech signal into a transcript in another language. 
Tracks are further organised into several specific tasks, covering specific languages. Training, development, and testing data for each track are freely available to the participants.

Potential participants in the evaluation are invited to check out our Call for Participation, fill in the Registration form, and join our e-mail list.

Permissible Training Data

MT Systems and Language Models for ASR:

Training of MT systems and language models for ASR is constrained to data supplied by the organizers or listed as permissible.

Participants can use any other linguistic resource provided that it does not include or exploits these TED and QED talks; any use of additional data with respect to that explicitly listed by the organizers has to be clearly stated in the system paper and communicated at the submission: such runs will be marked as "Unconstrained Training".

ASR Acoustic Modeling

As for ASR acoustic modeling no training data are distributed. Participants are allowed to use any publicly available data except for these TED and QED talks.