AI-powered enterprise web app that classifies elephant images as **African** or **Asian** using Deep Learning and Transfer Learning (MobileNetV2).
# 🐘 Elephant Species Classifier — Wild Sanctuary
> AI-powered enterprise web app that classifies elephant images as **African** or **Asian** using Deep Learning and Transfer Learning (MobileNetV2). Built with **FastAPI + Jinja2 + Tailwind CSS v3**.
---
## 🖼️ Screenshots
### 🏠 Home Page — Hero
### 🐘 Species Showcase — African & Asian Elephant Cards
### 🤖 AI Classifier — Upload Section
### 💛 Conservation CTA — Every Elephant Counts
### 📁 Image Upload — File Preview with Classify Button
### 🌏 Result Page — Asian Elephant Identified (68.98% Confidence)
### 🌍 Result Page — African Elephant Identified (87.09% Confidence)
### 🌍 Result Page — Conservation & Species Info Section
---
## ✨ Features
- 🔍 **Binary classification** — African 🌍 vs Asian 🌏 elephant
- 🧠 **MobileNetV2** pretrained on ImageNet, fine-tuned with transfer learning
- 🎨 **Enterprise dark UI** — drag-and-drop upload, animated confidence ring, glassmorphism cards
- ⚡ **FastAPI** backend with async request handling
- 🖼️ **Jinja2** server-side templates styled with **Tailwind CSS v3 (CDN)**
- 📊 Per-class confidence scores with animated progress bars & SVG donut chart
- 🌿 Full species profile page with habitat, traits, conservation status & fun facts
- 📱 **Fully responsive** — mobile hamburger menu, stacked layouts on small screens
- 🖼️ Asian elephant image gallery (3 AI-generated wildlife photos, served locally)
---
## 🗂️ Project Structure
```
Code/
├── main.py # FastAPI application
├── app.py # Legacy Streamlit app (reference)
├── requirements.txt # Python dependencies
├── class_indices.json # Label → index mapping
├── best_mobilenetv2.weights.h5 # Trained MobileNetV2 weights
├── elephant_model_xc.h5 # Trained Xception model (notebook)
├── EC (Final).ipynb # Training & evaluation notebook
├── demo/ # UI screenshots (1.png – 8.png)
├── templates/
│ ├── bas …