Video auditing model training method, video auditing method and related device

Abstract

The embodiment of the invention discloses a video auditing model training method, a video auditing method and a related device. The video auditing model training method comprises the following steps: acquiring a first sample image and a classification label of the first sample image; initializing a video auditing model, wherein the video auditing model comprises a primary sub-model and a secondary sub-model; training a first-level sub-model by adopting the first sample image, and calculating a classification loss rate of the first sample image classified by the first-level sub-model according to the classification label; and when the classification loss rate is greater than a preset value, training the secondary sub-model by using the first sample image. Since the first sample image whose classification loss rate is greater than the preset value is a difficult sample image whose positive and negative samples are difficult to distinguish, the difficult sample image can be adopted to train the secondary sub-model, so that the secondary sub-model learns the ability to distinguish the difficult samples, and finally the whole video auditing model can accurately distinguish the positive and negative samples. According to the invention, violation images in the video can be accurately determined, and the video submission accuracy is improved.

Description

technical field [0001] The embodiments of the present invention relate to the technical field of video review, and in particular, to a video review model training method, a video review method and related devices. Background technique [0002] With the explosive growth of the mobile Internet and the implementation of the Internet Security Law, content platform operators are facing more severe challenges. On the one hand, the increase in malicious users is on the one hand, and on the other hand, the supervision of illegal content in videos has been strengthened. Video content review can help enterprises Screen illegal images, videos, texts and other content in the platform, and through video content review, the illegal content can be filtered and deleted, so as to build a green and safe network environment for users. [0003] With the application of machine learning technology, videos are usually reviewed through trained video review models in the prior art. However, the live...

AI Technical Summary

Problems solved by technology

[0004] Embodiments of the present invention provide a video review model training method, video review method, device, electronic equipment, and storage medium to solve the problem that the video review model in the prior art is difficult to distinguish positive samples from negative samples, resulting in low review accuracy question

Images