Smart Product Comparison for Inventory Management
CatalogMatch is a desktop application that helps businesses compare new products against their existing inventory using AI-powered visual similarity analysis. Make smarter purchasing and inventory decisions by finding visually similar items in your catalog.
Visit our website: https://g1mliii.github.io/image-match/
- Visual Similarity Matching - Compare products based on color, shape, and texture
- Category Filtering - Match products only within the same category
- Batch Processing - Process up to 100 products at once
- Offline Desktop App - No internet required, your data stays private
- CSV Import/Export - Easy integration with your existing systems
- Detailed Scoring - See similarity breakdowns by feature type
- Side-by-Side Comparison - Visual comparison of matched products
- Download the app from our website
- Install by running the executable (Windows only)
- Upload your historical product catalog
- Match new products against your catalog
- Review results and export to CSV
- INSTALLATION.md - Complete installation guide for all platforms
- START_SERVER_README.md - How to start the server
- QUICK_REFERENCE.md - Quick reference guide
- Website Documentation - Online user guide
- GPU Setup Guide - GPU acceleration setup (AMD/NVIDIA/Intel/Apple Silicon)
- CLIP Developer Guide - Technical guide for CLIP integration
- Database Design - Database schema and design
- Matching Service - Matching algorithm details
- Windows 10 or later (64-bit)
- Python 3.12 (Required - Must be version < 3.13. Python 3.13+ is not compatible)
- 4 GB RAM minimum (8 GB recommended)
- 500 MB free disk space (1 GB with CLIP model)
- Display resolution: 1280x720 or higher
- AMD GPU: Radeon RX 6000/7000/9000 series + Python 3.12 + ROCm 6.4
- NVIDIA GPU: GeForce/RTX series (any recent GPU)
- Intel GPU: Arc/Iris Xe/UHD (11th gen+) + Intel Extension
- Apple Silicon: M1/M2/M3/M4/M5 (built-in MPS support)
Performance:
- AMD/NVIDIA GPU: 150-300 img/s (fastest)
- Intel GPU: 30-80 img/s (3-5x faster than CPU)
- CPU only: 5-20 img/s (slowest but works everywhere)
Setup GPU: See INSTALLATION.md or gpu/GPU_SETUP_GUIDE.md
Quick GPU Setup:
cd gpu
python setup_gpu.py- Free: Up to 50 products
- Pro: $29 one-time payment for unlimited products
Full documentation is available at: https://g1mliii.github.io/image-match/docs.html
- Backend: Python, Flask, OpenCV, NumPy
- Frontend: HTML, CSS (Tailwind), JavaScript
- Database: SQLite
- Desktop: PyWebView
- Visual similarity Powered by CLIP
- Category-based filtering system
- Batch upload and processing
- CSV metadata import/export
- Local SQLite database
- Desktop GUI application
- 100% Offline - All processing happens on your computer
- No Cloud Storage - Your data never leaves your machine
- Local Database - SQLite database stored in your app data folder
- No Tracking - We don't collect any usage data
- Documentation: https://g1mliii.github.io/image-match/docs.html
- Bug Reports: https://github.com/g1mliii/image-match/issues
- Business Inquiries: info@anchored.site
- macOS support
- Price history tracking
- Advanced filtering options
- Duplicate detection reports
- GPU acceleration
- API access for Pro users
Proprietary software. See Terms of Service for details.
- Free Version: Personal use only, up to 50 products
- Pro Version: Commercial use allowed, unlimited products
CatalogMatch is developed to help businesses make smarter inventory decisions through visual product comparison. Whether you're managing e-commerce inventory, retail stock, or product catalogs, CatalogMatch helps you quickly identify similar items and avoid duplicate purchases.
© 2024 CatalogMatch. All rights reserved.