🖼️ Smart Image Renamer

A powerful tool that lets you automatically rename image files based on their content, using a pretrained image classification model. It integrates seamlessly with the Windows context menu for quick access.

---

📑 Table of Contents

• Problem Statement
• Goal
• Features
• Future Improvements
• Demo
• Installation
• Usage
• Project Structure
• Documentation
• Technical Details
• Workflow Overview
• Tech Stack
• Model Requirements
• Contributing
• Roadmap
• License
• Author

---

✅ Problem Statement

When images are downloaded from the web, they are often named arbitrarily (e.g., image123.jpg, random123.png), making them hard to search or organize.

🎯 Goal

Create a Windows desktop application that:

• Takes an image as input.
• Uses a lightweight open-source ML model to analyze image content.
• Suggests a meaningful filename based on the image (e.g., sunset-beach.jpg, cat-sitting-sofa.jpg).
• Saves or renames the file with the generated name.
• Embed into Windows Explorer Right-Click Menu (Context Menu)
• Run as background service monitoring "Downloads" folder.

Inspiration

This Meme (Aug'25):

---

✨ Features

• 🔍 Uses AI to identify image content and rename accordingly.
• 🖱️ Right-click menu integration for easy access from Windows Explorer.
• 🛠️ Offline support - no internet needed to rename images.

---

✨ Future Improvements

• Decrease Latency for model indetifying the Image Tags
• Use a better, more lighteweight Image Captioning Model

---

🎥 Demo

---

⚙ Installation

To install Smart Image Renamer:

1. Download the latest ImageRenamerSetup.exe from Releases.
2. Run the installer. It will:
   • Copy required files to the install directory.
   • Register right-click context menu entries for image files.
3. Use "Uninstall Image Renamer" from the Start Menu to remove.

Note: Requires Python 3.9+ if running from source.

---

🚀 Usage

After installation, just right-click any .jpg, .png, .jpeg, etc., and select:

🖼️ Smart Rename with AI

To run from source:

python image_renamer.py path\to\images\

---

🗂 Project Structure

image-renamer/
├── image_renamer.py            # Main logic
├── installer.iss               # Inno Setup script
├── SmartRenameInstaller.exe    # Installer file by Inno Setup
├── build.bat                   # PyInstaller + EXE creation
├── README.md                   # This file
└── assets/                     # Icons and demo images

---

📚 Documentation

• Model: ResNet18 pretrained on ImageNet (from torchvision.models)
• Prediction flow: image → preprocessing → model → label → renaming
• Installer: built with Inno Setup, adds right-click handlers via registry
• EXE Packaging: via pyinstaller --onefile --add-data

---

🧠 Technical Details

This project is a Python-based file renaming utility that uses image classification to auto-suggest tags for images and rename them accordingly.

🏗 Workflow Overview

1. User Action:

   • Right-click an image → Select “Rename with Tags”.
   • Tool launches silently or in CLI (depending on build), takes image path as input.

2. Image Processing:

   • The image is passed through a lightweight image classification model (torchvision.models.squeezenet1_1).
   • The model returns the top 5 predicted class labels using pretrained weights on ImageNet.

3. Renaming Logic:

   • The original filename is updated with the predicted tags.
   • New format: originalfilename-tag1-tag2-tag3.jpg.

4. Packaging:

   • The project is bundled into a standalone .exe using pyinstaller.
   • The ImageNet labels file (imagenet_classes.txt) is embedded into the executable using the --add-data option.
   • An installer is generated using Inno Setup, with optional uninstaller support.

5. Silent Logging:

   • Logs of all renamed files and errors are stored in a logs/ directory next to the executable.

📦 Tech Stack

• Language: Python 3.11+
• ML Model: torchvision.models.squeezenet1_1
• Packaging Tool: PyInstaller
• Installer: Inno Setup Script (.iss)
• Silent Execution: Supports CLI invocation without UI

🧠 Model Requirements

• Small in size (can be bundled with the app, works offline).
• Fast inference on CPU (no GPU dependency).
• Decent image captioning ability (for naming).

---

🤝 Contributing

Contributions are welcome!

1. Fork the repo.
2. Create a branch: git checkout -b feature-name
3. Commit changes with clear messages.
4. Submit a PR with description and screenshots if relevant.

---

🛣 Roadmap

• Add support for JPG, PNG, WEBP, BMP
• Drag & drop UI
• Batch renaming preview window
• Mac/Linux versions
• Custom model support

---

📜 License

Distributed under the MIT License.

---

👤 Author

Karan Gupta