Gender recognition from images, some code implemented

Gender recognition from images, some code implemented


Job Description

I'm looking to offload some of my computer vision workload. Basically I need gender recognition from a facial image.

It has to be in python, so that I could use it with code such as this:

from LIB import GenderRecognition
import cv2
image = cv2.imread("lena2.jpg") # lets say this is just a 24x24 image of lena's face, cropped after detection with cascade
recognizer = GenderRecognition("Path_to_model")
gender = recognizer.getGender(image)

Gender can be just a probability between 0 and 1 where 0 is male and 1 is female.

I will also need the whole pipeline that trains the model, in case we will need to do some other stuff with the face.

Please give a bit of information regarding the process you will use or if you already have something like this done. So far I've coded a class that takes the images, detects the face, alligns the eyes and returns the cropped patch. If it would help to start from that point I'd gladly share the code. I'm currently using the LFW database.
So the next step is just the pre-processing, feature extraction (if any) and training. You can use any machine learning method you like; I was thinking of SVMs.

Also, please take into account that this is a price sensitive task. I'm willing to offer guidance if necessary, such as in the case of a student developer.

Other open jobs by this client