Video frame processing method and device, electronic device and storage medium

Abstract

The present application provides a video frame processing method and device, an electronic device and a storage medium, including: acquiring a plurality of key frame images of a video sample; extracting image features of each key frame image respectively to obtain a plurality of first image features; importing a plurality of first image features into an attention mechanism model and outputting a plurality of second image features based on the attention of an image channel; a plurality of second image features are fused to obtain fused features. The present application is based on an attentionmechanism, A method for capture associations between image of a plurality of key frames in a video sample based on image channels, and obtains the fusion feature through the feature fusion operation,In addition, the feature fusion operation does not limit the number of key frame frames, so that the processing of different time-length video can be realized, and the tedious operation degree of feature extraction for the video with irregular time-length is reduced.

Description

technical field [0001] The embodiments of the present application relate to the field of computer technology, and in particular, to a video frame processing method, device, electronic equipment, and storage medium. Background technique [0002] Recently, with the wide application of deep learning technology in multimedia sample classification processing and other related fields, many multimedia video sample classification functions have been developed for mobile terminal applications, and the information display and recommendation functions of applications have been optimized to improve user experience. . [0003] In related technologies, in a real scene, the first method is to classify videos, specifically by extracting a key frame of a video as a video cover, and based on the image feature extraction of the video cover, the classification features are obtained. Features can be used to label and classify videos. The second method is to extract multi-frame key frames from v...

AI Technical Summary

Problems solved by technology

[0005] Embodiments of the present application provide a video frame processing method, device, electronic equipment, and storage medium to solve the problem of information loss caused by a single key frame in the related art, resulting in a decrease in the classification accuracy of the video, and when extracting multiple key frames The problem of being unable to meet the high-efficiency demand for feature extraction of videos of variable duration

Images