Unlock instant, AI-driven research and patent intelligence for your innovation.
Facial expression recognition method and device based on CNN-Transform
What is Al technical title?
Al technical title is built by PatSnap Al team. It summarizes the technical point description of the patent document.
A technology of facial expressions and recognition methods, applied in neural learning methods, character and pattern recognition, computer components, etc., to achieve the effect of improving recognition accuracy and enhancing complementarity
Pending Publication Date: 2022-01-04
HOHAI UNIV
View PDF0 Cites 6 Cited by
Summary
Abstract
Description
Claims
Application Information
AI Technical Summary
This helps you quickly interpret patents by identifying the three key elements:
Problems solved by technology
Method used
Benefits of technology
Problems solved by technology
Although this method is more in line with human cognitive behavior, how to divide the image to better express facial features is a controversial topic
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more
Image
Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
Click on the blue label to locate the original text in one second.
Reading with bidirectional positioning of images and text.
[0054] The CNN-Transformer-based facial expression recognition method and device provided in this embodiment, its application process specifically involves the following steps:
[0055] S1, preprocess the input image, and obtain the corrected face image; the image comes from the RAF-DB dataset, the size of the image in the RAF-DB dataset is not fixed, and there are many interferences ot...
Embodiment 2
[0074] The present embodiment provides a kind of facial expression recognition device based on CNN-Transformer, comprising:
[0075] A preprocessing unit is used to preprocess the input image to obtain a corrected face image;
[0076] The local feature acquisition unit is used to calculate the LBP feature of the face image, and send it into the pre-built CNN network as input to obtain the local feature of the face;
[0077] The global feature acquisition unit is used to evenly divide the face image and send it to the Transformer to obtain the global feature of the face;
[0078] A fusion feature acquisition unit is used for information fusion of global features and local features to obtain fusion features;
[0079] The recognition result acquisition unit is used for performing emotion recognition by fusing features.
Embodiment 3
[0081] The present embodiment provides a CNN-Transformer-based facial expression recognition device, including a processor and a storage medium;
[0082] The storage medium is used to store instructions;
[0083] The processor is configured to operate according to the instructions to perform the steps of any one of the following methods:
[0084] Preprocess the input image to obtain the corrected face image;
[0085] Calculate the LBP feature of the face image, and send it as input to the pre-built CNN network to obtain the local features of the face;
[0086] Evenly divide the face image and send it to Transformer to obtain the global features of the face;
[0087] Information fusion is performed on global features and local features to obtain fusion features;
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More
PUM
Login to View More
Abstract
The invention discloses a facial expression recognition method and device based on CNN-Transform, and belongs to the technical field of digital imagesignalprocessing. The method comprises the steps: carrying out the preprocessing of an input image, and obtaining a corrected face image; calculating LBP features of the face image, and inputting the LBP features into a pre-constructed CNN network as input to obtain local features of the face; then uniformly dividing a face image, and sending the divided face image into a Transform to obtain global features of the face; and performing information fusion on the global features and the local features to obtain fusion features. and performing emotion recognition through feature fusion. According to the invention, through the feature fusion module, the influence weights of local features and global features on the whole can be autonomously learned; according to the method, complementarity among different features is effectively improved, so that emotion information contained in global features and local features can be fully utilized, and the recognition accuracy of emotion classification is improved.
Description
technical field [0001] The invention relates to a CNN-Transformer-based facial expression recognition method and device, belonging to the technical field of digital imagesignalprocessing. Background technique [0002] In daily communication, we can express emotions in both text and non-text forms, and the information that facial expressions can carry is especially important. In addition, facial expression recognition can be applied to the fields of fatigue driving detection, emotion synthesis, accurate advertisement placement, medical health and human-computer interaction, etc., and has broad application prospects. [0003] Facial expression recognition can be divided into traditional methods and deep learning methods. Traditional methods mainly extract manual shallow features such as local binary patterns (LBP), Gabor features, histograms of oriented gradients (HOG), bag-of-words features (BOW) and scale-invariant feature transforms (SIFT) as vector representations. The...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More
Application Information
Patent Timeline
Application Date:The date an application was filed.
Publication Date:The date a patent or application was officially published.
First Publication Date:The earliest publication date of a patent with the same application number.
Issue Date:Publication date of the patent grant document.
PCT Entry Date:The Entry date of PCT National Phase.
Estimated Expiry Date:The statutory expiry date of a patent right according to the Patent Law, and it is the longest term of protection that the patent right can achieve without the termination of the patent right due to other reasons(Term extension factor has been taken into account ).
Invalid Date:Actual expiry date is based on effective date or publication date of legal transaction data of invalid patent.
Login to View More 
Login to View More