Video classification model training method, device and equipment and storage medium

Abstract

The embodiment of the invention provides a video classification model training method, device and equipment and a storage medium, and relates to the technical field of video classification. The methodcomprises the steps: acquiring a label-free video set, wherein the label-free video set comprises at least one label-free video, and the label-free videos refer to videos which are not labeled manually; training a neural network model according to the label-free video set to obtain a trained neural network model, wherein the neural network model comprises a feature extraction module; cascading the trained feature extraction module with a randomly initialized video classification module to generate a video classification model; acquiring a label video set, wherein the label video set comprisesat least one label video, and the label video is a video marked manually; and training the video classification model through the label video set to obtain a trained video classification model. The embodiment of the invention can improve the training efficiency of the video classification model.

Description

technical field [0001] The embodiments of the present application relate to the technical field of video classification, and in particular to a training method, device, device and storage medium of a video classification model. Background technique [0002] The purpose of video classification is to label the input video with one or more labels containing information such as subject, scene, behavior, etc., such as: people, playground, running, etc. [0003] In related technologies, the video classification model can be used to complete the video classification. The video classification model needs to be trained before it is applied. Generally, the video classification model is trained through labeled videos. The video classification model requires a large-scale Labeled videos can make the trained video classification model more accurate. [0004] Since the prediction accuracy of the video classification model depends on the accuracy of the video label, although manual labeli...

AI Technical Summary

Problems solved by technology

[0004] Since the prediction accuracy of the video classification model depends on the accuracy of the video label, although manual labeling can improve the accuracy of the label, the efficiency of manual labeling is low, which leads to low training efficiency of the video classification model.

Images