To make any sensible commentary using LLMs, we must provide a good prompt with relevant stats either in tabular form or plain text. Thus, initialize a pipeline to extract these stats.
In order to understand the salient actions of a broadcast game, we introduce the pipeline of action spotting, which consists in finding all the actions occurring in the videos. Beyond soccer understanding, this task addresses the more general problem of retrieving moments with a specific semantic meaning in long untrimmed videos.
Targeted Output: {Pass, Drive, Header, High Pass, Out, Cross, Throw In, Shot, Ball Player Block, Player Successful Tackle, Free Kick, Goal}
These extracted data should also help in further data analysis of the game for coaches.
To make any sensible commentary using LLMs, we must provide a good prompt with relevant stats either in tabular form or plain text. Thus, initialize a pipeline to extract these stats.
In order to understand the salient actions of a broadcast game, we introduce the pipeline of action spotting, which consists in finding all the actions occurring in the videos. Beyond soccer understanding, this task addresses the more general problem of retrieving moments with a specific semantic meaning in long untrimmed videos.
Targeted Output: {Pass, Drive, Header, High Pass, Out, Cross, Throw In, Shot, Ball Player Block, Player Successful Tackle, Free Kick, Goal}
These extracted data should also help in further data analysis of the game for coaches.