Similar video screening method and device, equipment and storage medium

Abstract

The embodiment of the invention discloses a similar video screening method and device, equipment and a storage medium. The method comprises the following steps: constructing a similar video library ofeach to-be-selected video under a specified similar scale according to video-level characteristics of each to-be-selected video in a to-be-selected video library; searching similar videos of which the similarity with the to-be-selected video exceeds a preset similarity threshold from the similar video library of each to-be-selected video to obtain a corresponding video pair candidate library; andscreening out corresponding similar video pairs from the video pair candidate library based on the frame-level features of the two video parties of each candidate video pair in the video pair candidate library under the corresponding similarity weights. According to the technical scheme provided by the embodiment of the invention, the comprehensiveness and accuracy of similar video screening areimproved by adopting a double-screening mode, and meanwhile, the extraction operand of the frame-level features of the two video parties is controlled, so that the extraction operand of the frame-level features is reduced, and the screening efficiency of similar videos is improved.

Description

technical field [0001] The embodiments of the present invention relate to the technical field of video processing, and in particular to a method, device, device and storage medium for screening similar videos. Background technique [0002] With the rapid development of the short video and live broadcast industries, the growth scale of online videos continues to expand, and the daily increment of short videos has even reached tens of millions. At this time, the newly added videos with duplicate content have brought a lot of review redundancy to the review work. In addition, it also brings huge risks to the copyright protection work of video creators. Therefore, it is necessary to filter out similar videos with duplicate content to deduplicate newly added videos, so as to ensure the efficiency of video review and the protection of video creators. Copyright Security. [0003] At present, the similarity between different videos is usually calculated by extracting the features o...

AI Technical Summary

Problems solved by technology

At this time, for the features at the video-level granularity, each video only needs to perform a feature extraction step once to fuse the features of individual video frames under sparse sampling, and reduce the dimensionality to obtain the video-level granularity feature, which makes the feature dimension at the video-level granularity low, and cannot accurately represent the similarity between different videos; for the feature at the picture-level granularity, the spatial details of the video in different frames are considered, but for the video Multiple video frames in the video need to perform the feature extraction step, which makes the calculation of feature extraction too large, and splits the timing correlation of different video frames in a certain video, which greatly increases the similarity calculation efficiency between different videos.

Images