Real-Time Life Expectancy Prediction Using Deep Learning

Life Expectancy Prediction

Overview

The Life Expectancy Predictor is an advanced real-time web application that leverages deep learning to estimate an individual’s life expectancy. The application uses facial recognition through a webcam feed to detect faces, then predicts the person’s age and gender using pre-trained convolutional neural networks. Based on these predictions, it dynamically estimates life expectancy using a robust machine learning regression model.

This project combines computer vision, deep learning, and real-time video analysis with health data analytics, creating an intelligent system suitable for research, demonstrations, or future healthcare tools.

Project Information Table

Download New Real Time Projects :-Click here

Available Features

• Real-time Face Detection: Integrates OpenCV’s deep learning-based detector to identify and process faces from a live video stream.
• Age Prediction: Uses a pre-trained deep learning model (age_net.caffemodel) to determine the age range of the detected face.
• Gender Prediction: Utilizes a deep neural network (gender_net.caffemodel) to classify the gender as male or female.
• Life Expectancy Estimation: Applies a pre-trained RandomForestRegressor model to estimate the expected lifespan dynamically based on the predicted age and gender.
• Clean Interface Display: Displays results in real-time on screen with age, gender, and predicted life expectancy.

Technical Architecture and Implementation

1. Face Detection

The application uses OpenCV’s DNN module, relying on a Caffe-based model (res10_300x300_ssd_iter_140000.caffemodel) for detecting human faces from the webcam feed. It performs with high accuracy and low latency, making it suitable for live applications.

2. Age and Gender Estimation

Two separate CNN models are used for predicting age and gender:

• Age Model: Based on the age_net.caffemodel, this model outputs age buckets such as 0-2, 4-6, 8-12, etc.
• Gender Model: Trained on similar architecture using the gender_net.caffemodel, predicting gender as either “Male” or “Female”.

3. Life Expectancy Prediction

A machine learning regression model (life_expectancy_model_with_features.pkl) trained using RandomForestRegressor is responsible for the final estimation of life expectancy.

Model Training Details:

• Dataset: Life Expectancy Data sourced from Kaggle.
• Cleaning Steps:
  • Removed spaces in column headers.
  • Dropped rows with null values or imputed missing values with means.
• Feature Engineering:
  • Categorical values like “Status” (Developed/Developing) were encoded.
  • Non-informative columns like “Country” and “Year” were excluded.
• Model Performance: Evaluated with R² and Mean Squared Error (MSE), achieving reliable accuracy.

Requirements

To run this project, the following dependencies must be installed (as per requirements.txt in the ZIP):

opencv-python-headless==4.5.5.64
scikit-learn==0.24.2
pandas==1.3.3
joblib==1.0.1
numpy==1.21.2

How to Use

1. Install Dependencies: pip install -r requirements.txt
2. Run the Application: python main.py

The program opens a live video stream and begins displaying predictions instantly.

Professional Suitability

This project is structured for professional-level applications and demonstrates how deep learning can be effectively combined with real-world datasets to deliver actionable insights. It can serve as a foundational prototype for:

• Healthcare AI applications
• Research tools for demographic studies
• Smart kiosks or demo stations
• Real-time population analytics

No external database integration makes deployment simpler and faster, while the modular code allows for easy expansion or enhancement.

We have projects Available in all languages:–Click Here