A highly flexible interactive web-based tool for visualizing, editing, and managing YOLO format bounding box labels. Download and run the html file on your local system for 100% Privacy. No data leaves your server.
- Multiple Format Support:
.jpg,.jpeg,.png,.gif,.bmp,.webp,.nii.gz - Batch Processing: Upload and edit multiple image-label pairs simultaneously
- Navigation: Easy switching between image pairs with Previous/Next buttons
- High Resolution: Supports images of any size with automatic canvas scaling
- YOLO Format: Standard YOLO format (
class_id center_x center_y width height) - Real-time Interactive Editing: Edit either of coordinate values in text editor or bounding box corners on the image
- Custom Class Names: Map class IDs to human-readable names
- Visibility Toggles: Show/hide individual labels or all labels at once
- Automatic Pairing: Matches images and labels by filename (e.g.,
image.jpg↔image.txt) - Batch Upload: Select multiple files at once
- Custom Output Path: Specify a path + prefix for downloaded files or leave it empty for mainting the original name
- Save Current: Export currently edited labels
- Save All: Export all labels from all image pairs at once
- 100% Local Processing: All operations run in your browser
- No Server Required: No data transmission to any server
- Offline Capable: Works completely offline after initial page load
- No Installation: Pure HTML/JavaScript - no dependencies to install
-
Open the Application
- Click the launch link above or open
index.htmlin your browser - No installation or setup required
- Click the launch link above or open
-
Upload Files
- Click "Upload Multiple Image-Label Pairs"
- Select both image files and their corresponding
.txtlabel files - Files are automatically matched by basename (e.g.,
scan001.nii.gz+scan001.txt)
-
Start Editing
- Labels are initially hidden - toggle visibility as needed
- Click boxes to select them
- Drag to move or resize
Use Cases:
- Focus on specific objects by hiding others
- Reduce visual clutter
- Check individual label accuracy
┌─────────────────────────────────────────────────────┐
│ YOLO Label Editor │
│ Upload image-label pairs to begin │
├─────────────────────────────────────────────────────┤
│ [Upload Multiple Image-Label Pairs] │
├─────────────────────────────────────────────────────┤
│ [Previous] 1/10 - image001 [Next] Output: [path] │
│ [Save Current] [Save All] │
├─────────────────────────────────────────────────────┤
│ Class Names (Optional) │
│ ┌─────────────────────────────────────┐ │
│ │ person │ │
│ │ car │ │
│ │ ... │ │
│ └─────────────────────────────────────┘ │
├──────────────────────┬──────────────────────────────┤
│ Image Preview │ YOLO Labels │
│ ┌─────────────────┐ │ ┌─────────────────────────┐ │
│ │ │ │ │ 0 0.5 0.5 0.3 0.4 │ │
│ │ [Image with] │ │ │ 1 0.3 0.3 0.2 0.2 │ │
│ │ [Bounding] │ │ │ ... │ │
│ │ [Boxes] │ │ └─────────────────────────┘ │
│ │ │ │ │
│ └─────────────────┘ │ Label Visibility Controls: │
│ 0/3 visible │ ☐ Box 1 (person) [👁️] │
│ [Show All][Hide All]│ ☐ Box 2 (car) [👁️] │
│ │ ☐ Box 3 (dog) [👁️] │
└──────────────────────┴──────────────────────────────┘
- ✅ Chrome/Edge (recommended)
- ✅ Firefox
- ✅ Safari
- ✅ Opera
⚠️ IE 11 (not supported)
- Check that labels start hidden - click "Show All" button
- Verify YOLO format is correct (5 values per line)
- Ensure coordinates are between 0 and 1
- Ensure image and label files have identical basenames
- Example:
image.jpgmatchesimage.txt, notimage_labels.txt - File extensions:
.nii.gz→ remove.nii.gz,.jpg→ remove.jpg
- Make sure box is visible (eye icon on)
- Click to select box first (turns red)
- Drag from inside box to move, from handles to resize
- Check browser allows downloads
- Try different browser if issues persist
- File saves to default downloads folder
Contributions are welcome!
MIT License - feel free to use this tool for any purpose.
- Built with React and Tailwind CSS
- Inspired by the need for simple, privacy-focused labeling tools
- Thanks to the YOLO community
For issues, questions, or suggestions:
- Open an issue on GitHub
- Check existing issues for solutions
- Contribute improvements via pull requests
Potential future features:
- Undo/redo functionality
- Box creation (draw new boxes)
- Box deletion
- Zoom and pan for large images
- Export to other formats (COCO, Pascal VOC)
- Batch operations (delete all Class X)
Made with ❤️ for the computer vision community
⭐ Star this repo if you find it useful!