July 9-10, 2020 – Seattle, WA, USA

Call for Participation

The Evaluation Campaign of IWSLT 2020 will feature six challenges that address the following focus areas:

• Simultaneous speech translation (Chair: J. Gu, Facebook) – low latency, real-time interpretation
• Video speech translation (Chair: F. Huang, Alibaba) – multi-modal speech translation
• Offline speech translation (Chair: M. Turchi, FBK) – overcoming the cascaded approach
• Conversational speech translation (Chair: E. Salesky, JHU) – highly disfluent speech
• Open domain translation (Chair: A. Nagesh, Didi Labs) – new domains and languages
• Non-native speech translation (Chair: O. Bojar, Charles U.) – non-native disfluent speech

Training and development data for each challenge will be released. Organizers will provide baseline component systems to participants that do not have access to an individual component. Human and automatic scoring will be performed on each challenge, published and discussed.

Registration to the Evaluation Challenges is now open: Registration Form