A Fast Retrieval Technology of Face Database

Abstract

The invention relates to the field of artificial intelligence, and specifically provides a fast retrieval technology of a human face database. Compared with the prior art, the face feature vector output by the face recognition model of the present invention is a multidimensional vector, wherein each dimension is a floating-point numerical scalar, and the space formed by the face feature vector is considered to be R n space. It is characterized in that it includes the following steps: step 1, establishing an orderly segmentation of the face database to ensure that each divided sub-interval is in order; step 2, based on the divided sub-intervals, comparing the face feature vectors during face recognition , according to the distance between the feature vector and the center vector to judge the partition. According to the method of partitioning and storing the face records in the database in advance according to the distance, and properly indexing, the candidate records that need to be searched can be greatly reduced when recognizing faces, and the comparison speed can be increased.

Description

technical field [0001] The invention relates to the field of artificial intelligence, and specifically provides a fast retrieval technology of a human face database. Background technique [0002] With the development of artificial intelligence and deep learning technology, the deep neural network model has been widely used in face recognition tasks, and has greatly improved the accuracy of face recognition. The trained face recognition model calculates the face feature vectors of all personnel and stores them in the database. When identifying an employee, calculate the feature vector of the face to be recognized in the model calculation, and then calculate the vector The distance to all feature vectors in the database, if the minimum value of the distance is less than the preset threshold δ, it is judged as the same person; otherwise, it is judged that the person is not in the database. [0003] Under this method, many face recognition scenarios require real-time recognitio...

AI Technical Summary

Problems solved by technology

[0003] Under this method, many face recognition scenarios require real-time recognition performance, but when there are too many people in the face database, the process of finding the minimum distance will take too long, resulting in a decline in the real-time performance of the application, or even impossible. accept

Images