🏷️ YOLO Inlay Hints

YOLO Inlay Hints is a professional-grade visualization and productivity extension for VS Code, specifically engineered for the Computer Vision annotation workflow.

It bridges the gap between raw ML annotation data (Class IDs) and human-readable context, enabling instant validation, navigation, and debugging within complex annotation datasets.

Designed for high-performance environments, the extension ensures zero-latency feedback while working with large multi-folder YOLO datasets.

🎥 Visual Demo

Example:

0 0.512 0.433 0.231 0.123

Becomes:

0 [person] 0.512 0.433 0.231 0.123

🚀 Key Features

Contextual Labeling

Instantly maps cryptic Class IDs (e.g., 0) to human-readable labels like:

0 → [person]

Hints appear dynamically within valid YOLO annotation files.

Intelligent File Discovery

The extension automatically locates your mapping file (classes.txt) by:

• Searching upward from the current document
• Performing a workspace-wide parallel scan if necessary

This allows flexible dataset structures without manual configuration.

Jump-to-Definition (Navigation Provider)

Hold:

• Cmd + Click (Mac)
• Ctrl + Click (Windows / Linux)

on a Class ID to immediately open the mapping file at the exact definition line.

Contextual Hover & Metadata

Hover over any inlay hint to view metadata including:

• Class label
• Source mapping file
• Full path of the active mapping file

This helps confirm which dataset configuration is currently active.

Dataset Validation

If a Class ID is missing from the mapping file, the extension shows a Warning Inlay Hint, allowing instant debugging of annotation integrity.

Example:

⚠ Unknown Class ID

🛠️ Technical Architecture

This extension implements robust, production-grade engineering patterns suitable for professional ML pipelines.

Cross-Platform File System API

Uses the VS Code URI-based filesystem API (vscode.workspace.fs) enabling compatibility with:

• Local environments
• Remote SSH development
• Cloud VMs (Azure / AWS)
• Docker containers
• WSL
• GitHub Codespaces

Race Condition Mitigation

Implements Cancellation Tokens to safely cancel background tasks when VS Code requests termination.

This prevents:

• UI lag
• flickering hints
• unnecessary CPU usage during rapid scrolling or file switching

Asynchronous Disk I/O Caching

A Map-based caching layer stores file contents in memory, reducing disk reads and ensuring zero-latency editor interaction.

Content-Based Validation

A built-in schema validator ensures hints activate only in valid YOLO annotation files.

Required format:

ClassID X Y Width Height

Example:

0 0.521 0.433 0.212 0.123

⚙️ Extension Configuration

📦 Installation

Install directly from the VS Code Marketplace.

1. Open VS Code
2. Go to Extensions
3. Search for YOLO Inlay Hints
4. Click Install

📁 Usage

1. Open a folder containing YOLO annotation files
2. Ensure a classes.txt (or configured mapping file) exists
3. Open any annotation .txt file

Hints will appear automatically.

Example Dataset

dataset/
│
├─ images/
│   ├─ img1.jpg
│
├─ labels/
│   ├─ img1.txt
│
└─ classes.txt

🧠 Typical Workflow

1. Annotate images using a labeling tool
2. Open the dataset in VS Code
3. Instantly visualize class labels inside annotation files
4. Navigate between labels using Cmd/Ctrl + Click

🤝 Contributing

Contributions are welcome.

If you'd like to improve the extension:

1. Fork the repository
2. Create a feature branch
3. Submit a pull request

📄 License

MIT License