Deep network migration learning method facing a marked noise apparent age database

Abstract

The invention discloses a deep network migration learning method facing a marked noise apparent age database. The apparent age database is randomly divided into two parts according to the preset proportion, one is training set, and the other is verification set. The proportion of training set is higher than verification set. A small amount of data is randomly extracted from the training set and repeated n times to obtain n sub-training sets. The set of the remaining data in the training set is denoted as data set A. According to the transfer learning method, n classification models are obtained by in-depth learning of n sub-training sets, and then the data set A is identified by using n classification models. The deep network migration learning method facing the marked noise apparent age database enables the apparent age database with high accuracy to be obtained when the marked noise is weakened by the apparent age database, and can effectively weaken the influence of the marked noiseon the experimental result and make the result more credible.

Description

technical field [0001] The invention relates to a method for a marked noise apparent age database, in particular to a deep network migration learning method for a marked noise apparent age database. Background technique [0002] A person's biological age is calculated from the moment of birth, and one year is added to the biological age every time one year passes, which shows that the biological age will not be changed by the influence of the external environment. However, in the apparent age database, these apparent ages are related to a person's cultivation, responsibility, experience, and psychology, and each person's performance in these aspects is also different, so when a person's age When judging by appearance, there will be a gap with the biological age. Here, the database obtained by marking the age of a person according to the appearance of the human eye is called an apparent age database. In the process of labeling the apparent age data, one person may be used f...

AI Technical Summary

Problems solved by technology

If the apparent age database containing label noise is not processed to a certain extent, then if such an age database is used directly, the obtained results will contain certain errors and the accuracy will be low

[0003] There are many mature data processing technologies in the database field. However, for the above-mentioned problem of data noise caused by human subjective judgment factors, these technologies cannot solve this problem.

In addition, marker noise exists not only in the apparent age database, but also in other aspects of the data. There are still different degrees of noise in different forms. At present, there is no general and effective method to completely reduce the noise

Images