Exhibition

13Smart Production

Automatic Audio Description from Sports-Related Data for Live Broadcasting
Audio description service for live programs

Outline

We are researching “automatic audio description” for sports events. Audio descriptions currently target recorded programs such as dramas since their application to live broadcasting has been problematic. We here introduce technology for automatically describing the progression of a sports match by speech synthesis using sports-related data in live sport broadcasting.

サービスのイメージ

Features

Efficient production
Large-scale international sports events can span several days with multiple events held at the same time. The proposed audio descriptions are automatically generated from sports-related data, resulting in an efficient production process requiring no human intervention. The script generated in this process can also be used as captions without modification.
Flexible wording
How the sports-related data is to be read can be decided beforehand to achieve appropriate wording depending on context. Voice quality can also be selected as desired.
Supplementary broadcast service
For Rio Olympic and Paralympic Games, NHK delivered video and venue audio of nonbroadcast events over the Internet. The addition of automatic audio description would provide a service that makes such content even more comprehensible and enjoyable for many people.

Future plans

We are developing elemental technologies with a view to trials at 2018 PyeongChang Winter Olympic and Paralympic Games.

  • The research on speech synthesis technology is jointly conducted with AI Inc.
pageTop

Please enter the exhibition booth number.

X
J M T P
1 2 3
4 5 6
7 8 9
Del 0 Enter

Automatic Audio Description from Sports-Related Data for Live Broadcasting Audio description service for live programs

  • Automatic Audio Description from Sports-Related Data for Live Broadcasting
  • Flow of automatic audio description text