A sophisticated machine learning–driven web application that estimates and visualizes individual CO2 emissions using Convolutional Neural Networks (CNN).
Built with a focus on scalability, usability, and environmental sustainability.
- CNN-Based Prediction – Accurately estimates carbon footprint from user lifestyle inputs.
- Interactive Web Interface – Built with Streamlit for real-time input & visualization.
- Data Pipelines – Efficient preprocessing with Pandas and NumPy.
- Visual Insights – Category breakdowns and global comparisons via Matplotlib and Seaborn.
- Cloud-Ready Architecture – API-first design for easy deployment and scaling.
| Category | Tools & Libraries |
|---|---|
| Programming | Python |
| Machine Learning | TensorFlow, Keras (CNN) |
| Data Handling | Pandas, NumPy |
| Visualization | Matplotlib, Seaborn |
| Frontend | Streamlit |
| Deployment | Cloud-ready, API-based integration |
flowchart TD
A[User Input: Transport, Energy, Diet, Waste] --> B[Data Preprocessing - Pandas, NumPy]
B --> C[CNN Model - TensorFlow/Keras]
C --> D[Emission Prediction - CO2 tonnes per year]
D --> E[Visualization - Matplotlib, Seaborn]
E --> F[Streamlit Dashboard - Real Time Insights]
F --> G[Country Comparison and Sustainability Analytics]
Usage
1.Enter lifestyle inputs (transport, electricity, diet, waste).
2.Submit for CNN-powered carbon footprint prediction.
3.Explore interactive dashboards with category-wise breakdowns.
4.Compare results against global and national averages.
Future Enhancements
1.Cloud Deployment on AWS / GCP / Streamlit Cloud.
2.API Integration for modular sustainability apps.
3 Advanced ML Models (RNN, ensembles, time-series forecasting).
4.Personalized Recommendations for reducing emissions.
- Global Data Integration (IPCC, OWID).
Author
Priyam Parashar
Biotechnology Engineering Student | Software, AI & Sustainability Enthusiast
📍 VR Academy, Bengaluru
🎓 Mentor: Mr. Vijayanand