Non-standard state face age estimation method based on feature aggregation
The invention provides a non-standard state face age estimation method based on feature aggregation, and relates to the technical field of biological feature recognition. Face alignment and illumination normalization are carried out on the image, and illumination enhancement is carried out on the im...
Saved in:
Main Authors | , , , |
---|---|
Format | Patent |
Language | Chinese English |
Published |
13.08.2024
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | The invention provides a non-standard state face age estimation method based on feature aggregation, and relates to the technical field of biological feature recognition. Face alignment and illumination normalization are carried out on the image, and illumination enhancement is carried out on the image with low brightness by using an MSRCR algorithm; based on the residual convolutional neural network, extracting age features of the face and calculating a top-5 prediction set of each image; a node of a feature corresponding graph of an image and the number of elements in a prediction set intersection between every two nodes are used as weights of edges to form a graph, then feature aggregation is carried out based on a graph attention network, internal correlation between features is captured, and finally node classification is carried out. According to the method, the accuracy of face age estimation is improved, and the method has extremely high application value in the aspect of juvenile network addiction pr |
---|---|
Bibliography: | Application Number: CN202410597428 |