Multimedia content evaluation method and device and training method thereof

Abstract

The invention discloses a training method of a multimedia content evaluation model, and the evaluation model comprises an evaluation network module and a post-processing unit; the evaluation network module comprises a classification network sub-module and a subsequent dimension reduction sub-module, and the training method comprises the following steps: training the evaluation network module; inputting the plurality of picture data into the trained evaluation network module to obtain evaluation network module output, and inputting the evaluation network module output into a post-processing unit for processing to determine post-processing parameters for converting the evaluation network module output into a final evaluation value. The invention also provides a multimedia content evaluationmethod and evaluation device, and electronic equipment. According to the scheme, the training data acquisition difficulty can be effectively reduced, learning of multimedia content features is betterdriven, and therefore the automation degree, efficiency and accuracy of multimedia content evaluation are improved.

Description

technical field [0001] The present application relates to the technical field of multimedia content processing, in particular to a multimedia content evaluation method and device. The present application also relates to a training method for a multimedia content evaluation model. In addition, the present application also relates to related electronic equipment. Background technique [0002] Currently, multimedia content including User Generated Content (UGC), such as music, short videos or videos, is generated in large quantities and widely disseminated. [0003] However, a large amount of multimedia content, especially user-generated content, is uneven, which hinders better and faster mining of high-quality content. [0004] For example, in terms of multimedia content, such as short video review, the number of daily short video platforms has reached 100,000. If a purely manual audit is used, it will not only take a long time but also cost a lot of labor. For a large num...

AI Technical Summary

Problems solved by technology

However, there is a lack of effective means to recommend high-quality content to new users who lack behavioral data

Interest-based recommendations may also cause some newly generated high-quality content to be effectively pushed to platform users

This has a negative impact on multimedia platforms to extend user retention and attract new users

Images