Facial shape tracking method based on space-time cascade shape regression

Abstract

The present invention discloses a facial shape tracking method based on space-time cascade shape regression. The method comprises: detecting the first frame of a video through combination of a face detector configured to detect and register, initializing the whole system, and obtaining five facial feature points; assessing the five facial feature points to obtain similar transformation parameters (rotation, offset and scale) and face poses (left sides of the faces, right sides of the faces and the faces); and employing multi-view cascade shape regression to predicate the face shape of the current frame, when the registering result confidence is larger than a setting threshold, allowing the time sequence regression to set about tracking the face shapes, and when the registering result confidence is smaller than a setting threshold, starting a re-initialization mechanism to perform stable tracking of the face shapes. The facial shape tracking method based on space-time cascade shape regression is faster in convergence speed and higher in precision through the multi-view (the left side of the face, the right side of the face and the face) cascade regression, and faster in the face shape tracking speed and more accurate in the face shape tracking through the time sequence regression and the re-initialization mechanism.

Description

technical field [0001] The invention relates to a face shape tracking method based on space-time cascade shape regression, and belongs to the fields of computer vision technology and multimedia technology. Background technique [0002] Registration in video is the basis of video analysis. After accurate registration of faces in video, facial expression analysis, face recognition, face pose estimation, and individual behavior identification can be performed. At present, face registration based on static images has achieved good results, and the speed and accuracy have been greatly improved. However, when the existing algorithms are directly applied to videos, there are still many challenges, mainly from video sequences. Changes in facial expression, illumination, occlusion, and posture. Compared with the face registration of a single image, the face registration in the video needs to fully consider the relationship between two adjacent frames. If the registration deviation o...

AI Technical Summary

Problems solved by technology

At present, face registration based on static images has achieved good results, and the speed and accuracy have been greatly improved. However, when the existing algorithms are directly applied to videos, there are still many challenges, mainly from video sequences. Changes in facial expression, illumination, occlusion, and posture

Compared with the face registration of a single image, the face registration in the video needs to fully consider the relationship between two adjacent frames. If the registration deviation of the face in the previous frame is large, the next frame will The registration of the face causes errors, and the errors will accumulate, which will eventually lead to the failure of the face registration in the video.

Images