Build an Accurate Age and Gender Detection Model Using Python

Is it possible to determine someone's age and gender based solely on facial features? Yes. This project uses computer vision and pre-trained deep learning models to illustrate precisely that. In this project, we create an age and gender detection model that uses a facial image as input and makes the following predictions:

• The individual's gender (male or female)
• The age range, for example, 0–2, 4–6, 8–12, etc.

We will employ OpenCV's DNN module with pre-trained Caffe models to make predictions using a small sample set of facial images. There is no need for manual training. Real-world uses for this technique include interactive systems, surveillance, and targeted advertising.

See our Top 25+ Essential Data Science Projects GitHub to Explore in 2025 blog for more project ideas like this one.  

What Should You Know Beforehand?

It is better to have at least some background in:

• Python programming – functions, loops, and libraries, such as NumPy & OpenCV.
• Image processing concepts – how pixels, channels, and image arrays work.
• Working with Google Colab – uploading files, installing packages, and running notebooks

Technologies and Libraries Used

For this project, the following tools and libraries will be used:

Pretrained Models Used for Prediction

Below are the pretrained models that we will be utilizing:

Time Taken and Difficulty Level

On average, it will take about 1 to 2 hours to complete. Duration may increase/decrease depending on your familiarity with - Python, image processing, & ML concepts. It’s best for beginner to intermediate level.

How to Build an Age and Gender Detection Model

Let’s start building the project from scratch. We will start by:

• Clone a sample facial image dataset from a GitHub repository
• Load and explore these images for testing purposes
• Use OpenCV to detect faces and prepare images for model input (resizing, normalization)
• Load pre-trained Caffe models for gender and age prediction using OpenCV's DNN module
• Run predictions on new images and display the results with bounding boxes and labels

Without any further delay, let’s start!

Step 1: Clone the Dataset from GitHub

Let’s clone the dataset from the public GitHub repository. This dataset contains thousands of labeled facial images organized into gender and age categories.

Here is the code to do so:

# Clone the dataset repository from GitHub
!git clone https://github.com/smahesh29/Gender-and-Age-Detection.git

# Navigate into the directory
%cd Gender-and-Age-Detection

Output:

Cloning into 'Gender-and-Age-Detection'...

remote: Enumerating objects: 72, done.

remote: Counting objects: 100% (25/25), done.

remote: Compressing objects: 100% (11/11), done.

remote: Total 72 (delta 20), reused 14 (delta 14), pack-reused 47 (from 1)

Receiving objects: 100% (72/72), 90.73 MiB | 17.76 MiB/s, done.

Resolving deltas: 100% (25/25), done.

/content/Gender-and-Age-Detection

The output tells us that - The GitHub repo has been successfully cloned and is now ready to be accessed in the Colab environment.

Step 2: Load the Pre-trained Models

In this step, we will load the Caffe-based deep learning models for Age and Gender Detection. The models are already trained and stored as .caffemodel and .prototxt files.

Use the below-mentioned code:

import cv2

# Define paths for model files
gender_model = 'gender_net.caffemodel'
gender_proto = 'gender_deploy.prototxt'
age_model = 'age_net.caffemodel'
age_proto = 'age_deploy.prototxt'

# Load the models
gender_net = cv2.dnn.readNet(gender_model, gender_proto)
age_net = cv2.dnn.readNet(age_model, age_proto)

Step 3: Define Label Lists and Image Preprocessing

Now, we will define the age and gender categories. Besides this, we will also set up a utility function to preprocess input images for the models.

Use the code given below to do so:

# Mean values used during model training
MODEL_MEAN_VALUES = (78.4263377603, 87.7689143744, 114.895847746)

# Define class labels
gender_list = ['Male', 'Female']
age_list = ['(0-2)', '(4-6)', '(8-12)', '(15-20)', 
            '(25-32)', '(38-43)', '(48-53)', '(60-100)']

# Image preprocessing function
def preprocess_image(image_path):
    img = cv2.imread(image_path)
    face_img = cv2.resize(img, (227, 227))  # Resize to model input size
    blob = cv2.dnn.blobFromImage(face_img, 1.0, (227, 227), MODEL_MEAN_VALUES, swapRB=False)
    return blob, img

Step 4: Load Models and Predict Age & Gender

In this step, we will - 

• Load the pre-trained models.
• Predict age and gender.
• Display the result on the image.

Use the below-mentioned code to accomplish the same:

import cv2
import matplotlib.pyplot as plt

# Load the models
age_net = cv2.dnn.readNetFromCaffe(
    "/content/Gender-and-Age-Detection/age_deploy.prototxt",
    "/content/Gender-and-Age-Detection/age_net.caffemodel"
)
gender_net = cv2.dnn.readNetFromCaffe(
    "/content/Gender-and-Age-Detection/gender_deploy.prototxt",
    "/content/Gender-and-Age-Detection/gender_net.caffemodel"
)

# Define age and gender lists
AGE_LIST = ['(0-2)', '(4-6)', '(8-12)', '(15-20)',
            '(25-32)', '(38-43)', '(48-53)', '(60-100)']
GENDER_LIST = ['Male', 'Female']
MODEL_MEAN_VALUES = (78.4263377603, 87.7689143744, 114.895847746)

# Preprocess function
def preprocess_image(image_path):
    img = cv2.imread(image_path)
    if img is None:
        raise ValueError(f"Image not found: {image_path}")
    face_img = cv2.resize(img, (227, 227))
    blob = cv2.dnn.blobFromImage(face_img, 1.0, (227, 227), MODEL_MEAN_VALUES, swapRB=False)
    return blob, img

# Run prediction
image_path = "/content/Gender-and-Age-Detection/man1.jpg"  # Change to another image if needed
blob, original_img = preprocess_image(image_path)

# Predict gender
gender_net.setInput(blob)
gender_pred = gender_net.forward()
gender = GENDER_LIST[gender_pred[0].argmax()]

# Predict age
age_net.setInput(blob)
age_pred = age_net.forward()
age = AGE_LIST[age_pred[0].argmax()]

# Annotate image
label = f"{gender}, {age}"
cv2.putText(original_img, label, (10, 30),
            cv2.FONT_HERSHEY_SIMPLEX, 0.8, (0, 255, 0), 2)
cv2.rectangle(original_img, (5, 5), (220, 40), (0, 255, 0), 1)

# Convert BGR to RGB for displaying
rgb_img = cv2.cvtColor(original_img, cv2.COLOR_BGR2RGB)

# Show image
plt.figure(figsize=(6, 6))
plt.imshow(rgb_img)
plt.axis('off')
plt.title("Prediction Result")
plt.show()

Output:

Conclusion

We predicted age and gender from photos using OpenCV and pre-trained Caffe models. After processing a face, the model produced the following result: Male, 60–100 years old. The system predicts one set of labels per image, despite being trained to classify faces into two gender categories and eight age groups. For ease of interpretation, the results were superimposed on the picture. This demonstrates that the model has a respectable level of accuracy in identifying and categorizing facial features.

Unlock the power of data with our popular Data Science courses, designed to make you proficient in analytics, machine learning, and big data!

Explore our Popular Data Science Courses

Elevate your career by learning essential Data Science skills such as statistical modeling, big data processing, predictive analytics, and SQL!

Top Data Science Skills to Learn

Stay informed and inspired with our popular Data Science articles, offering expert insights, trends, and practical tips for aspiring data professionals!

Read our popular Data Science Articles

Reference:
https://colab.research.google.com/drive/15M_eSIfZ1QxhUSzROhMTJYtlHRufaG6U?usp=sharing