========== Face Attribute Recognition ===========

Objective :
***********
	 The main objective of this system is to predict the different types of facial attributes 
	 such as(eyes: open/closed, Smiling: yes/No, Glasses: Wearing/Not Weaaring, Gender and Age)
	 from the given input image.

Process :
**********
	 Here we used many trained models on the face images and the Models we used are as follows.
	 	Models :
	 	-------------
	 	1. get_frontal_face_detector() and shape_predictor_68_face_landmarks.dat are the two 
	 	   pre_trained models used in detecting the faces and 68 landmarks on face using those
	 	   two models.

	 	   		The above two models are used in predicting wether the eye is closed or opened by 
	 	   	calculating the eucledian distances between the 6 landmark points on each eye. From 
	 	   	this we can calculate the 'eye_aspect_ratio' and by using the threshold value of 0.2
	 	   	we can predict wether the eye is opened or closed.

	 	2. We trained the model on the faces with "happiness" and "neutral" images from the "fer13
	 	   dataset" which predicts wether the person is smiling or not.

	 	3. We trained the model on the faces with "glasses" and "without glasses" dataset and we 
	 	   are using the model in predicting the person is wearing glasses or not.

	 	4. Here we are using a model which is trained on predicting the gender and age of the 
	 	   person from others resource .

	 The above models are used in predicting the face atributes.
	 	   We have used the gray scale images with predicted faces from the get_frontal_face-detector
	 	   and we cropped those images and used them in predicting the eyes, smile and glasses.
	 	   Then we used the cropped image and increased borders by 20% and here we used colored 
	 	   new_cropped images in predicting the age and gender.

Applications:
*************
	 1. It can be further be used in generating the captions/sentences based on given image.
	 2. Further it can be used in the Facial Expression/Emotion detection (wether he is angry, 
	 	happy, fear, etc.)
	 3. Can be used in detecting wether the person is sleeping or awake while driving cars to
	 	avoid car accidents.