Rapid classification based on image using residual neural networks

Abstract

Every human being is created with a certain gender, female or male. Gender differences provide opportunities for mating and producing offspring for regeneration. With the development of the world of information technology, the recognition of gender based on digital photos can be done for the needs of biodata, permission to access public toilets and others. The process of sex classification based on photographs with the use of residual neural networks has been carried out in this study. The process consists of several stages, namely the learning process of the features of each image that has been classified in a male or female class. The next process was carried out by resnet classification of 3,354 pictures (jpg) of men (1414 files) and women (1940 files). The data divided into 2 parts, 80% for training, 20% for testing data. The results of total images of 588 from total available data obtained an accuracy rate of 89.49%.