IWSLT 2020 - Third Call for Participation

[Marcello Federico]

The International Conference on Spoken Language Translation

ACL – 17th IWSLT 2020 – Third Call for Participation

July 9-10, 2020 – Seattle, WA, USA

http://iwslt.org

The International Conference on Spoken Language Translation (IWSLT) is an annual scientific conference 

for the study, development and evaluation of spoken language translation technology: speech-to-text, 

speech-to-speech translation, simultaneous and consecutive translation, speech dubbing, cross-lingual 

communication including all multimodal, emotional, para-linguistic, and stylistic aspects and their applications 

in the field.  The conference organizes evaluations and presents scientific work and system descriptions.  

Due to the coronavirus pandemic, ACL-IWSLT 2020 will either be postponed or organized as a virtual (online) 

conference. Accepted paper will be published in the ACL proceedings.

IMPORTANT DATES

Evaluation Campaign:

April 20, 2020: end of evaluation period

April 24, 2020: system description deadline

May 11 - 2020: review feedback

May 18 - 2020: camera-ready version due

Scientific program:

April 13, 2020: paper submission deadline
May 11, 2020: paper notification
May 18, 2020: camera-ready version due

 
EVALUATION CHALLENGES

IWSLT 2020 will feature six challenges that address the following focus areas:
●      Simultaneous translation (Chair: J. Gu, Facebook) – low latency, real-time translation
●      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
 
 
CONFERENCE TOPICS

IWSLT invites submissions of scientific papers to be published in the ACL proceedings and presented either in oral or poster format. 

The conference selects high quality, original contributions on theoretical and practical issues of automatic speech recognition (ASR), 

machine translation (MT) speech synthesis (TTS) pertaining to spoken language translation (SLT).  More specific topics include:
 
Approaches and performance
Under-resourced languages
End-to-end models
Multimodal Interfaces
Evaluation, metrics, benchmarks, usability
Architectures
Paralinguistic aspects in interpretation
Adaptation
Language resources
Post- and pre-processing
Open source software
Efficiency, speed, latency
Multilingual approaches
Sign language translation
Applications
Discourse and context aware translation