A Synchronous Recognition Method of Face Identity and Expression

Abstract

The present invention proposes a method for synchronous recognition of face identity and expression. First, facial feature extraction is performed on each face image, and corresponding semantic features are defined for each image at the same time, and kernel principal component analysis (PCCA) is used for facial features. ) feature fusion method, so that the input image features have better recognition characteristics. On this basis, the partial least squares regression (PLsR) method is used to establish the relationship model between facial features and semantic features, and this model is used to recognize the expression and identity of the face image to be recognized. Experiments show that the method proposed by the invention can not only realize the simultaneous recognition of human face and expression, but also improve the recognition rate of human facial expression recognition.

Description

technical field

[0001] The invention relates to a face recognition method, in particular to a synchronous recognition method of face identity and expression. Background technique

[0002] Facial images contain rich information, and through facial images, not only people's identity, but also their facial expressions can be recognized. At present, facial expression recognition and identity recognition have become two hot research issues in the fields of computer vision and pattern recognition. The main goal of facial expression recognition is to extract the main features that can reflect emotional categories from facial images, and then perform expression classification and recognition on this basis. Most traditional facial expression recognition methods classify facial images into one of seven basic expression types (happy, sad, surprised, angry, disgusted, scared, neutral). Similar to the facial expression recognition method, the goal of face recognition is to match an unk...

Images