Wedding AI Photo Selector

An AI-powered computer vision system that automatically selects the best wedding photos from thousands of images.

This project helps wedding photographers and studios reduce photo selection time from several hours to just a few minutes.

---

Overview

Wedding photographers typically capture 3000–10000 photos per event. Manually selecting the best images for albums is time-consuming.

This AI system automatically analyzes the entire photo set and selects the best images for album design and delivery.

---

Features

Image Quality Detection

Detects and removes blurry or low-quality images.

Duplicate Image Removal

Identifies and removes duplicate or burst-shot images.

Face Detection

Detects faces in photos to prioritize images with people.

Bride & Groom Priority Detection

AI prioritizes images containing the bride and groom.

AI Aesthetic Scoring

Each image receives an aesthetic score based on visual quality.

Scene Classification

Images are automatically categorized into scenes such as:

• Bride portrait
• Groom portrait
• Couple portrait
• Wedding ceremony
• Family photos
• Stage photos
• Emotional moments
• Dance celebrations
• Candid moments

Story-Based Photo Selection

Groups images into a timeline of the wedding story.

Best Image Ranking

Ranks images and selects the top photos for album creation.

---

AI Pipeline

RAW IMAGES
     ↓
Image Loading
     ↓
Blur Detection
     ↓
Duplicate Removal
     ↓
Face Detection
     ↓
Bride/Groom Detection
     ↓
AI Aesthetic Scoring
     ↓
Scene Classification
     ↓
Story Builder
     ↓
Image Ranking
     ↓
Best 500 Album Photos

---

Project Structure

wedding-ai-photo-selector
│
├── app.py
├── main.py
├── requirements.txt
│
├── src
│   ├── pipeline.py
│   ├── duplicate_remover.py
│   ├── face_detector.py
│   ├── scene_classifier.py
│   ├── story_classifier.py
│   │
│   └── features
│       ├── aesthetic_scorer.py
│       └── quality_checker.py
│
├── data
│   ├── demo_images
│   └── selected_images
│
└── architecture
    ├── pipeline.png
    └── screenshots

---

Installation

Clone the repository

git clone https://github.com/sundhars2823/Machine-learning-protfolio.git

Navigate to the project

cd wedding-ai-photo-selector

Create virtual environment

python -m venv venv

Activate environment (Windows)

venv\Scripts\activate

Install dependencies

pip install -r requirements.txt

---

Run AI Photo Selection Pipeline

python main.py

The system will process all images and automatically select the best photos.

---

Run Web Interface

This project includes a web interface built with Streamlit.

Run the app:

streamlit run app.py

Open in browser:

http://localhost:8501

---

Technologies Used

Python OpenCV PyTorch OpenCLIP NumPy Streamlit

---

Performance

Typical Wedding Dataset

Input images: 4000 – 8000 photos Processing time: 5 – 15 minutes Output: 300 – 500 best photos

Manual photo selection time reduced by 90%.

---

Future Improvements

• Automatic wedding album layout generator
• Emotion detection
• Couple moment prioritization
• Lightroom plugin
• Automatic album export for printing
• AI slideshow generator

---

Author

SUNDHARESHWAR ASPIRING ML ENGINEER

---

License

This project is for research and educational purposes.