Why Opencv predict the gender as male most of the time [closed]
I did follow the steps in the Opencv docs, But the result most of the time is male !!!
I'm using Logitech HD Pro Webcam C920 .
Yale database is a challenging database due to illumination conditions. Illumination normalization is necessary to cope with these conditions. But I don't think this is the best database to train your classifier. You should train with google images, for example. In http://docs.opencv.org/trunk/modules/contrib/doc/facerec/tutorial/facerec_gender_classification.html the database has 8 male and 5 female, each with 10 images.