Face image pose estimation and correction method and system, medium and electronic equipment

Abstract

The invention provides a face image pose estimation and correction method and system, a medium and electronic equipment. The method comprises the following steps: collecting face image data, labeling face key points, calculating a rotation and translation vector label of a face, and establishing a training data set and a test data set; setting a rotation and translation vector loss function, training a constructed convolutional neural network model by using the training data set, and stopping training until the loss calculated by the rotation and translation vector loss function reaches a preset threshold value; and evaluating the trained convolutional neural network models by using the test data set, and selecting an optimal convolutional neural network model as a target network model. According to the face image pose estimation and correction method and system, the medium and the electronic equipment, face pose estimation can be completed, a face image can be corrected and aligned through affine transformation, and the adverse effect of a face detector correction and alignment method on subsequent attribute recognition and feature extraction module application is removed.

Description

technical field [0001] The present invention relates to the technical field of face image processing, in particular to a face image pose estimation and correction method, system, medium and electronic equipment. Background technique [0002] Face pose estimation and correction alignment play an important role in the fields of face attribute classification and face feature recognition. Face pose estimation is to estimate the pose of the face in the 2D image in the 3D world coordinate system. , face correction alignment is to correct and align the faces with incorrect angles in the two-dimensional image. Faces with large face angles (such as side faces, head up, head down, etc.) will lead to an increase in the distance between attribute classes and face feature vectors, thereby affecting the accuracy of the corresponding model. [0003] At present, face attribute classification and face feature recognition are widely used in intelligent business, video surveillance and other ...

AI Technical Summary

Problems solved by technology

[0005] In view of the shortcomings of the prior art described above, the object of the present invention is to provide a method, system, medium and electronic equipment for face image pose estimation and correction To solve the problem of low efficiency of face pose estimation and correction in the prior art

Images