theger14/fyp-ma-13

A study on the factors that affect gender bias in classification

fyp-ma-13's Issues

Implement functional gender classification models

Previously, we had designed some preliminary models as an initial prototype during the Project Design phase.

Now, implement improved versions of that model with and without perturbations (such as makeup) for further testing.

Design and implement a graphical user interface

A graphical user interface for image classification would be ideal for our project to demonstrate our findings.

Implement an application with the following features:

• Loads any JPG or PNG image from the file system
• Loads a single TensorFlow-compatible image classification model
• Performs gender classification for the image
• Displays the relevant statistics for the image and the classification model

Run experiments

on glasses & make up data alone - 10%, 25% and 50%