AI-Powered Photo Management & Face Recognition Desktop Application

Organize • Detect • Recognize • Cluster • Explore

Overview • Features • Technology • Screenshots • Installation • Usage

SmartPhotoVision is an intelligent desktop photo management application designed to make large photo collections easier to organize, analyze, and explore.

The application combines Artificial Intelligence, Computer Vision, Face Detection, Face Recognition, Face Clustering, and SQLite database management into a single desktop interface.

Instead of manually searching through hundreds or thousands of images, SmartPhotoVision analyzes photographs and helps organize them according to the people detected in the images.

• 📂 Scans photo folders recursively
• 👤 Detects human faces using YuNet
• 🧠 Generates face embeddings using SFace
• 🧩 Groups similar faces into personas
• 🗃️ Stores photo and face information in SQLite
• 🖼️ Supports multiple common image formats
• 📍 Handles photo metadata such as date and GPS information
• ⚡ Processes photo collections in the background
• 🖥️ Provides a modern desktop interface using PySide6
• 🔄 Supports face reclustering for improved organization

SmartPhotoVision follows a multi-stage analysis pipeline:

                    📁 Photo Folder
                          │
                          ▼
                🔍 Recursive Scanning
                          │
                          ▼
                  🖼️ Image Loading
                          │
                          ▼
                 👤 YuNet Detection
                          │
                          ▼
                 🧠 SFace Embedding
                          │
                          ▼
                 🔎 Face Comparison
                          │
                          ▼
                  🧩 Face Clustering
                          │
                          ▼
                    👥 Personas
                          │
                          ▼
                 🗃️ SQLite Storage
                          │
                          ▼
                   🖥️ Desktop GUI

The project uses two OpenCV ONNX models:

Face Detection

└── face_detection_yunet_2023mar.onnx

Used to locate faces inside photographs.

models/ └── face_recognition_sface_2021dec.onnx

Used to generate face embeddings and compare detected faces.

SmartPhotoVision/
│
├── 📄 main.py
│
├── 🧠 analyzer.py
│
├── 🗃️ database.py
│
├── 🛠️ paths.py
│
├── 🤖 models/
│   ├── face_detection_yunet_2023mar.onnx
│   └── face_recognition_sface_2021dec.onnx
│
├── 🖥️ gui/
│   ├── window.py
│   └── theme.py
│
└── 📘 README.md

---

Component Responsibility main.py Application startup and configuration analyzer.py Photo scanning, face detection, recognition and clustering database.py SQLite database and data persistence paths.py Application and resource path management gui/ Desktop user interface and visual styling models/ AI models used for face analysis

🐍 Python 🖥️ PySide6 👁️ OpenCV 🔢 NumPy 🗃️ SQLite Artificial Intelligence 🤖 YuNet Face Detector 🧠 SFace Face Recognition 🧩 Face Embeddings 🔎 Similarity Matching 📊 Face Clustering

Visual Studio Code Git GitHub Python Virtual Environment

🧠 AI 👤 Face Analysis 🗃️ Storage 🖥️ Interface YuNet + SFace Detection + Recognition SQLite PySide6

🏷️ Project SmartPhotoVision 📦 Application Type Desktop Application 🐍 Language Python 🖥️ GUI Framework PySide6 👁️ Computer Vision OpenCV 🗃️ Database SQLite 🤖 AI Models YuNet + SFace 🔢 Version 1.0.4 📸 Screenshots

Real application screenshots are provided below as proof of the working output.

Main SmartPhotoVision desktop interface.

Photo scanning and AI-powered face analysis.

Detected faces within the analyzed photographs.

Detected faces organized into personas based on similarity.

Organized photo collection inside the application.

To display your actual screenshots on GitHub, create:

screenshots/
│
├── main-window.png
├── photo-analysis.png
├── face-detection.png
├── personas.png
└── photo-library.png

Then replace the example images above with your actual screenshots.

1️⃣ Clone the Repository git clone https://github.com/your-username/SmartPhotoVision.git cd SmartPhotoVision

2️⃣ Create a Virtual Environment Windows python -m venv venv venv\Scripts\activate Linux / macOS python3 -m venv venv source venv/bin/activate

3️⃣ Install Dependencies pip install -r requirements.txt

▶️ Run the Application

Start SmartPhotoVision with:

python main.py

The application will initialize the database and launch the desktop interface.

Step 1 — Launch

Run:

python main.py Step 2 — Select Photo Folder

Choose the directory containing your photographs.

Step 3 — Analyze

Start the photo analysis process.

SmartPhotoVision will:

📂 Scan Images
      ↓
👤 Detect Faces
      ↓
🧠 Generate Embeddings
      ↓
🔎 Compare Faces
      ↓
🧩 Cluster Similar Faces
      ↓
🗃️ Save Results
Step 4 — Explore

Use the application interface to explore the analyzed photographs and detected personas.

SmartPhotoVision uses SQLite to maintain persistent application data.

Main Tables
photos
│
├── Photo information
├── Filename
├── Dimensions
├── Modification time
├── Taken-at metadata
└── GPS information

persons
│
├── Person ID
└── Person Name

faces
│
├── Photo reference
├── Person reference
├── Bounding box
├── Detection score
├── Face embedding
└── Pin status

tags
│
└── Photo categorization

photo_tags
│
└── Photo ↔ Tag relationships

SmartPhotoVision is designed as a local desktop application.

Photo analysis and application data are handled locally through the application's processing and SQLite storage components.

No cloud photo-upload workflow is required for the core analysis pipeline.

Always review the project's actual configuration and dependencies before using it with sensitive photo collections.

The application is designed to process photo collections efficiently by:

🔄 Processing images in the background
📐 Resizing very large images before analysis
🧠 Limiting stored embeddings per persona
🧩 Using similarity-based clustering
💾 Persisting analysis results in SQLite
🛑 Supporting analysis cancellation
📝 Providing logging and error handling
📁 Supported Image Formats

SmartPhotoVision supports:

.jpg
.jpeg
.png
.webp
.bmp
.tif
.tiff

Face detection and recognition are integrated directly into the application workflow.

OpenCV provides the underlying computer-vision capabilities for detecting and analyzing faces.

SQLite provides structured persistence for photographs, detected faces, personas, and tags.

PySide6 provides the graphical interface and application experience.

Face similarity and clustering allow photographs to be organized around detected people.

Potential future development areas include:

🌐 Multi-language interface
📱 Companion mobile application
🔎 Advanced semantic photo search
🏷️ Automatic image tagging
🧠 Improved face clustering
👥 Better person management
📊 Collection analytics dashboard
⚡ GPU acceleration
📤 Export and backup functionality
🎯 Advanced search and filtering
🤝 Contributing

Contributions, suggestions, and improvements are welcome.

Contribution Workflow
Fork
  ↓
Create Branch
  ↓
Make Changes
  ↓
Test
  ↓
Commit
  ↓
Pull Request

Please ensure that proposed changes are tested before submitting a pull request.

Add your preferred open-source license here.

For example:

MIT License

If you find SmartPhotoVision useful or interesting:

MS Computer Science Student Specialization in Artificial Intelligence

Institute of Management Sciences, Peshawar, Pakistan

I enjoy building complete software systems—from networking and operating system concepts to backend architecture, machine learning, and distributed systems.

• 🎓 Degree / Field: MS in Computer Science (Specialization in Artificial Intelligence)
• 🏆 Achievements: Designed and executed 30+ machine learning and data visualization projects delivering actionable insights
• 🎯 Focus Areas: Data Visualization, AI/ML, Big Data Mining
• 📬 Contact: shahidazam2020@gmail.com
• WhatsApp: Chat on WhatsApp

For the README above, I recommend organizing the repository like this:

SmartPhotoVision/
│
├── 📄 main.py
├── 📄 analyzer.py
├── 📄 database.py
├── 📄 paths.py
├── 📄 requirements.txt
│
├── 🤖 models/
│   ├── face_detection_yunet_2023mar.onnx
│   └── face_recognition_sface_2021dec.onnx
│
├── 🖥️ gui/
│   ├── window.py
│   └── theme.py
│
├── 📸 screenshots/
│   ├── main-window.png
│   ├── photo-analysis.png
│   ├── face-detection.png
│   ├── personas.png
│   └── photo-library.png
│
└── 📘 README.md

If you find this repository useful for learning Machine Learning, consider giving it a ⭐.

Your feedback, suggestions, and contributions are welcome.

🧠 Learn Machine Learning • 💻 Build Models • 📊 Analyze Data • 🚀 Create Projects

Made with ❤️ for learning, experimentation, and continuous improvement.