Instantly replay any scene you want while watching a sports broadcast

This is the professional baseball season that every sportsman has been waiting for. Those who can’t go to the stadium are comforted by watching 스포츠중계 at home. The team you are cheering for is about to score an important goal! ‘Tarreung- Follow-Long-‘ the phone rings. I missed an important moment while answering the phone.

In general, when watching sports broadcasts, there are scenes that you miss or want to see again. However, the video provided in the broadcast cannot be selected and viewed by the viewer. Sometimes I watch the scenes I want to skip over and over again, and sometimes I wait until the scene I want comes out. In addition, since the video uploaded to the main portal site after the game is over shows only the main game scenes, viewers may not be able to find the scene they want.

The ETRI research team succeeded in developing a broadcasting platform technology that automatically creates and replays video of major events based on broadcast video and text message of sports such as baseball and basketball. It is the so-called ‘viewer-customized broadcasting platform technology’. It’s like a magic technique that takes you back in time to a past scene. Viewers can select and view event types such as shots or home runs that they want to see again in sports broadcasts such as baseball and basketball, or major scenes by team and player.

More fun with your own highlight video!

ETRI researchers have automated the match to provide real-time footage. A video similar to the edited result of hand-made work is provided within a few tens of seconds. Viewers can classify the play scenes of a specific player or team they want, so they can provide their own highlight video.

Viewer-customized broadcasting platform technology includes event information acquisition technology using text relay, video information extraction technology through deep learning, and automatic event section analysis technology.

It is the principle of finding the relevant event section by recognizing the screen composition or movement of players in the video relay by acquiring information about the occurrence of events such as shoot, rebound, hit, and home run in the relay video using text relay. To this end, the computer automatically classifies and extracts information such as pitching motion, infield, outfield, game time, and number of pitches through deep learning technology.

The extracted video information analyzes the relay video production pattern and introduces a machine learning algorithm to identify the event section. After that, event section information is recorded by matching which section of the relay video the event identified through text relay corresponds to, and collected as a video.

Customized broadcasting, aiming to provide ‘my own replay’ service

The viewer-customized broadcasting platform technology is expected to be used for services that automatically classify images of interest to viewers during sports broadcasting in real-time streaming or digital broadcasting in the future, and can be viewed again within tens of seconds. In the case of IPTV operators, this technology can be applied to set-top boxes to provide customized replay services to service users. In addition, it was developed to be linked with mobile devices so that you can receive the event service you want without interfering with TV viewing, and archiving service providers will be able to automatically index a vast amount of video and provide it to users. The ETRI research team plans to expand the field of broadcasting not only baseball and basketball, but also various sports with a consistent video production pattern, to form a business alliance or technology transfer with IPTV operators, relay video archiving service providers, and clubs that require tactical analysis. Meanwhile, the viewer-customized broadcasting platform technology was exhibited at the International Broadcasting Equipment Show (NAB) and received great acclaim, such as receiving attention from FOX Sports, an NBA broadcaster.

Leave a Reply

Your email address will not be published. Required fields are marked *