Lip synthesis video verification method, device, computer equipment and storage medium

Abstract

The embodiment of the invention discloses a mouth shape synthetic video verification method and device, computer equipment and a storage medium. The method comprises the following steps: acquiring character material video data, and determining a current frame video according to the character material video data; performing face detection on the current frame of video to obtain a face detection result; judging whether the human face detection result is a human face and the human face is not shielded or not; if so, intercepting a face area of the current frame of video; performing mouth detection on the face area to obtain a mouth detection result; judging whether the mouth detection result is a result that the mouth is in the face area or not; if not, determining that the current frame video is a problem frame video, and storing the serial number of the problem frame video in a problem frame serial number list; judging whether the current frame of video is the last frame of video of the character material video data or not; and if not, intercepting the next frame of video to form the current frame of video. By implementing the method provided by the embodiment of the invention, the number of missed detection problem frames can be reduced, and the working efficiency is improved.

Description

technical field [0001] The present invention relates to a computer, and more particularly to a lip-synched video verification method, device, computer equipment and storage medium. Background technique [0002] With the exponential growth of audiovisual content consumption, the rapid creation of video content has become a typical requirement. Lip-synthesis technology allows the characters in the video to "speak" the audio content through the character video material and audio. The video effect of lip-synch synthesis depends on the provided video material. When the face area, especially the mouth area, is occluded, it will seriously affect the synthesized video effect. Therefore, the video material needs to be checked and verified before synthesis. [0003] The longer the audio duration corresponding to the lip synthesis, the longer the corresponding video duration. For a video of the same length, the higher the video frame rate, the more total frames and the more frames pe...

AI Technical Summary

Problems solved by technology

For a video of the same length, the higher the video frame rate, the more total frames, and the more frames per second, the smoother the displayed action will be, and the higher the similarity between frames will be. The technology is to find problematic frames through manual verification, find suspected problematic frames, and need to check repeatedly to determine which consecutive frames are problematic frames, because the appearance of problematic frames is random, and it is necessary to manually search for problematic frames for a long time Checking will lead to visual fatigue, decreased concentration, and an increase in the number of missed problem frames, which will eventually lead to a decrease in work efficiency

Images