Automatic action recognition method based on match video

Abstract

The invention belongs to the technical field of network data transmission, and particularly relates to an action recognition method based on a match video. The method comprises the following steps: preprocessing and screening event videos; carrying out keyword labeling and statistics; carrying out action and semantic feature extraction; and performing association training learning on the action and semantic features. According to the action recognition method based on the match video, the match rebroadcasting cost can be reduced, the requirements for personnel and equipment are reduced, a basic model structure based on contents such as machine language automatic recognition and semantic matching is provided for contents such as match rebroadcasting live broadcast, and a technical basis isprovided for realizing competition video-based action recognition for competition rebroadcasting or live broadcasting.

Description

technical field [0001] The present application belongs to the technical field of network data transmission, and in particular relates to an automatic action recognition method based on a game video. Background technique [0002] With the improvement of network video data transmission capabilities and the continuous development of various sports among people of all ages, people's demand for on-site and video viewing of various sports events continues to increase. In the process of each competition, there is often more than one competition. In addition to the complete video recording and broadcasting of the main competition, the allocation of software and hardware resources for sub-competitions or secondary competitions is not the same, often not Equipped with a complete broadcast support and commentary team, but there are corresponding fan teams to pay attention to some sub-competition and sub-competition events, so it is impossible to give them good broadcast and commentary ...

AI Technical Summary

Problems solved by technology

[0002] With the improvement of network video data transmission capabilities and the continuous development of various sports among people of all ages, people's demand for on-site and video viewing of various sports events continues to increase. In the process of each competition, there is often more than one competition. In addition to the complete video recording and broadcasting of the main competition, the allocation of software and hardware resources for sub-competitions or secondary competitions is not the same, often not Equipped with a complete broadcast support and commentary team, but there are corresponding fan teams to pay attention to some sub-competition and sub-competition events, so it is impossible to give them good broadcast and commentary support for some fans. On the other hand, the semantic automatic recognition technology for machine AI is also developing continuously, providing a technical basis for the automation of on-site broadcasting and explanation. Currently, there are various simple task descriptions and linear prompts on the production line based on this technology. applied, but there is no complete technical solution for the automatic broadcast of events

Images