Imsearch

Imsearch is a desktop application for searching, comparing, and discovering similar images within a directory. Built with a custom-themed, user-friendly interface using CustomTkinter, Imsearch leverages deep learning models to analyze and rank images by similarity, making it ideal for photographers, designers, researchers, and anyone working with large image collections.

Key Features

• Intuitive UI: Clean, responsive interface with dark/light theme toggle.
• Easy Image Upload: Quickly upload an image to use as a search query.
• Directory Selection: Choose any folder to search for similar images.
• Deep/Nested Search: Option to search subdirectories for images.
• Similarity Sorting: Images are ranked from most similar to least similar using advanced models (DINO and MobileNet).
• Progress Feedback: Real-time progress bar and loading spinner for long searches.
• Model Selection: Automatic selection between DINO (for accuracy) and MobileNet (for speed).
• Result Visualization: Grid view of results with image previews; double-click to open images directly.

How It Works

1. Upload an Image: Select an image as your search query.
2. Select a Directory: Choose the folder containing images to search.
3. Choose Search Options: Enable "Deep Check" to include subfolders.
4. Start Search: Click "Check" to begin. The app will process and rank images by similarity.
5. View Results: See the most similar images in a grid, sorted by relevance. Double-click any image to open it.

Demo

🎬 Imsearch Demo Video

Download Application

📁 Imsearch.zip

How to build

To build imsearch, follow these steps:

1. Clone the repository:

2. Set Up the Environment

Install dependencies using Conda (recommended):

3. Run the Application

Model Files

• DINO and MobileNet model files are required in the models/ directory. These are used for feature extraction and similarity computation.