Video processing method, device, computer equipment and readable storage medium

By introducing a multi-head visual attention module into the video classification model, the keyframe features are processed, and the problems of low accuracy and high computational volume in the prior art are solved, thereby achieving higher video classification accuracy and lower computing costs.

CN113963302BActive Publication Date: 2025-03-28BEIJING WEIBOYI TECH CO LTD
2 Cites 0 Cited by

Patent Information

Application Number
CN202111325596.9
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2021-11-10
Publication Date
2025-03-28
Estimated Expiration
2041-11-10

AI Technical Summary

Technical Problem

It is difficult to design a video classification model with high accuracy in the prior art, especially when processing video timing information and high redundant data, the calculation is large and the model is easy to overfit.

Method used

A video classification model built on the multi-head visual attention module is used to process keyframe features to improve the accuracy of video classification. The multi-head visual attention module can obtain richer feature information and prevent model overfitting.

Benefits of technology

Through the multi-head visual attention module, the accuracy of video classification is significantly improved, the risk of model overfitting is reduced, and the calculation amount and parameter amount are optimized.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The present application proposes a video processing method, device, computer equipment and readable storage medium, which relates to the field of image processing technology. The method proposed in the present application includes: obtaining key frame features corresponding to at least one key frame image of a video to be classified; processing the key frame features corresponding to at least one key frame image through a pre-trained video classification model to obtain a classification result of the video to be classified, wherein the pre-trained video classification model is constructed based on a multi-head visual attention module. The solution can be applied in fields such as computer vision and artificial intelligence.
Need to check novelty before this filing date? Find Prior Art