AgeGenderEthnicityPrediction

This project leverages machine learning techniques and computer vision to predict a person's age, gender, and ethnicity from a given image. Through this project, I had the chance to use the tensorflow library for the first time in python. In the content of the project, I developed a project that predicts the age, gender and ethnicity of the people in the pictures with the data we get from the dataset on kaggle. Apart from these, you can see certain tables using data visualization algorithms related to certain data.

Dataset Link (Kaggle): https://www.kaggle.com/code/shahraizanwar/age-gender-ethnicity-prediction/data

import

import numpy as np
import pandas as pd
import tensorflow as tf
import tensorflow.keras.layers as L
import matplotlib.pyplot as plt
import plotly.graph_objects as go
import plotly.express as px
from sklearn.model_selection import train_test_split

Distributions

## normalizing pixels data
data['pixels'] = data['pixels'].apply(lambda x: x/255)

## calculating distributions
age_dist = data['age'].value_counts()
ethnicity_dist = data['ethnicity'].value_counts()
gender_dist = data['gender'].value_counts().rename(index={0:'Male',1:'Female'})

def ditribution_plot(x,y,name):
    fig = go.Figure([
        go.Bar(x=x, y=y)
    ])

    fig.update_layout(title_text=name)
    fig.show()

Age Distribtion

ditribution_plot(x=age_dist.index, y=age_dist.values, name='Age Distribution')

Ethnicity Distribution

ditribution_plot(x=ethnicity_dist.index, y=ethnicity_dist.values, name='Ethnicity Distribution')

Gender Distribution

ditribution_plot(x=gender_dist.index, y=gender_dist.values, name='Gender Distribution')

Sample Images

plt.figure(figsize=(16,16))
for i in range(1500,1520):
    plt.subplot(5,5,(i%25)+1)
    plt.xticks([])
    plt.yticks([])
    plt.grid(False)
    plt.imshow(data['pixels'].iloc[i].reshape(48,48))
    plt.xlabel(
        "Age:"+str(data['age'].iloc[i])+
        "  Ethnicity:"+str(data['ethnicity'].iloc[i])+
        "  Gender:"+ str(data['gender'].iloc[i])
    )
plt.show()

Model for Gender Prediction

y = data['gender']

X_train, X_test, y_train, y_test = train_test_split(
    X, y, test_size=0.22, random_state=37
)

Build and train model

model = tf.keras.Sequential([
    L.InputLayer(input_shape=(48,48,1)),
    L.Conv2D(32, (3, 3), activation='relu', input_shape=(32, 32, 3)),
    L.BatchNormalization(),
    L.MaxPooling2D((2, 2)),
    L.Conv2D(64, (3, 3), activation='relu'),
    L.MaxPooling2D((2, 2)),
    L.Flatten(),
    L.Dense(64, activation='relu'),
    L.Dropout(rate=0.5),
    L.Dense(1, activation='sigmoid')
])

model.compile(optimizer='sgd',
              loss=tf.keras.losses.BinaryCrossentropy(),
              metrics=['accuracy'])


## Stop training when validation loss reach 0.2700
class myCallback(tf.keras.callbacks.Callback):
    def on_epoch_end(self, epoch, logs={}):
        if(logs.get('val_loss')<0.2700):
            print("\nReached 0.2700 val_loss so cancelling training!")
            self.model.stop_training = True

callback = myCallback()

model.summary()

history = model.fit(
    X_train, y_train, epochs=20, validation_split=0.1, batch_size=64, callbacks=[callback]
)

fig = px.line(
    history.history, y=['loss', 'val_loss'],
    labels={'index': 'epoch', 'value': 'loss'},
    title='Training History')
fig.show()

---------------------------------------
loss, acc = model.evaluate(X_test,y_test,verbose=0)
print('Test loss: {}'.format(loss))
print('Test Accuracy: {}'.format(acc))