Gender Detection Using CNN model

Project Platform: Dphi - Gender Determination by Morphometry of Eyes

Project Duration: July, 2021 - Aug, 2021

Project Description:

• Problem Statement

The anthropometric analysis of the human face is an essential study for performing craniofacial plastic and reconstructive surgeries. Facial anthropometrics are affected by various factors such as age, gender, ethnicity, socioeconomic status, environment, and region.

Plastic surgeons who undertake the repair and reconstruction of facial deformities find the anatomical dimensions of the facial structures useful for their surgeries. These dimensions are a result of the Physical or Facial appearance of an individual. Along with factors like culture, personality, ethnic background, age; eye appearance and symmetry contributes majorly to the facial appearance or aesthetics.

Our objective is to build a model to scan the image of an eye of a patient and find if the gender of the patient is male or female.

• Solution:

Solution proposed was a CNN model which will predict the gender, it takes image input size of 100x100x3, consisting 3 convolution layers and 9 hidden layers, which was trained on nearly 2300 images which gave accuracy of around 89.5% on test set. Optimizer used in the model is "ADAM optimizer" which adjusts learning rate value while training the model so that model is not over fitted on data set.

Skills:

• Neural Networks ; Convolutional Neural Networks

• Deep Learning