Video splitting method, device and equipment

Abstract

The present application discloses a video splitting method, device and equipment, which are used in the field of video processing and relate to cloud technology and artificial intelligence technology. Determine the video unit sequence corresponding to the video to be processed, perform feature extraction on each video unit in the video unit sequence, and obtain a feature sequence including initial video features corresponding to each video unit. Each video unit is regarded as a target video unit, and the target video features of the target video unit are determined according to the similarity between the target video unit and other video units and their corresponding initial video features, so that the features of the video unit have global modeling capabilities. According to the characteristics of the target video, it is predicted whether the target video unit belongs to the split boundary, and the prediction result sequence corresponding to the feature sequence is obtained. According to the prediction result sequence, the video to be processed is split to avoid splitting using local information and artificial design criteria, which solves the problem. The split result is susceptible to the problem of local noise.

Description

technical field [0001] The present application relates to the field of video processing, in particular to a video splitting method, device and equipment. Background technique [0002] In recent years, with the popularity of various video websites and short video applications (Application, APP), people's demand for high-quality video content is also increasing, and video creators' demand for high-quality video editing tools is also gradually increasing . As an important part of video clipping tools, splitting a video into multiple segments through a video splitting algorithm has attracted more and more attention. [0003] The current video splitting algorithms are generally implemented based on artificial intelligence, and often rely on local modeling and artificial strategies to complete. This method is easily affected by local noise, and different artificial strategies need to be designed to adapt to various Application scenarios affect the practicability and robustness o...

AI Technical Summary

Problems solved by technology

[0004] In order to solve the above technical problems, this application provides a video splitting method, device, equipment and storage medium, so that the characteristics of the video unit have the ability of global modeling, avoiding the use of local information for judgment, and then relying on a large number of manual designs Criteria, thus solving the problem that the video splitting results are easily affected by local noise, and improving the accuracy of video splitting

Images