Japanese-->English Video Guided MT Task for Ambiguous Subtitles

[HOME]

INTRODUCTION

We present the video guided ambiguous subtitle translation task for Japanese --> English translation.

TASK DESCRIPTION

This is a Japanese to English multimodal machine translation task. We provide VISA, an ambiguous subtitles dataset, for training, validation and testing. The dataset contains parallel subtitles in which the Japanese source subtitles are ambiguous and may require visual information in corresponding video clips for disambiguation. Furthermore, according to the cause of ambiguity, the dataset is divided into Polysemy and Omission. Participants can use the standard training data to train their multimodal machine translation system.

Corpora

• The VISA dataset can be downloaded from here here. You can find more information about VISA dataset in this page.

Submission Details

• Constrained submission: Using the VISA dataset only.
• Unconstrained submission: Using pre-trained models, additional data from other sources. (Please contact organizers before using anything)

CONTACT

For general questions, comments, etc. please email to "wat-organizer -at- googlegroups -dot- com". For questions related to this task contact "prajdabre -at- gmail -dot- com" or "chu -at- i -dot- kyoto-u -dot- ac -dot- jp".

NICT (National Institute of Information and Communications Technology)
Kyoto University
Last Modified: 2021-01-04