YOLIC Labeling Tool

The YOLIC Labeling Tool is a user-friendly annotation software designed to support the You-Only-Look-at-Interested-Cells (YOLIC) methodology for real-time object detection based on cell-wise segmentation.

Unlike traditional object detection methods that require scanning the entire image (leading to high computational cost), YOLIC focuses only on predefined Interested Cells (COIs). This enables efficient detection and classification using a single lightweight network — ideal for edge devices.

⚠️ Important: The YOLIC Labeling Tool is designed for annotating datasets to train YOLIC models, not for direct object recognition.

---

✨ Key Features

• Flexible cell configuration system
  Import custom detection areas based on task requirements.

• Polygon-based annotation
  Precise object localization within interested cells.

• Semi-automatic labeling
  Integrates deep learning models to accelerate annotation.

• RGB and RGB-D support
  Works with both standard and depth-enhanced images.

• Customizable granularity
  Control the level of detail in object representation.

---

🚀 Getting Started

You can install and use the YOLIC Labeling Tool in two ways:

• Option 1: Install directly using the precompiled .msi installer (recommended for most users)
• Option 2: Build the software from source code (recommended for developers)

---

🔹 Option 1: Install Using the MSI Installer (Recommended)

1. Go to the Releases section of the repository.
2. Download the Tool.msi file.
3. Run the installer on your Windows machine.
4. Follow the installation instructions.

For detailed usage instructions, please refer to the Usage.pdf file included in the repository.

🧪 Test the Tool

You can test the annotation tool using:

• test image.zip
• Configuration3.json

These files provide a ready-to-use example configuration.

---

🔹 Option 2: Build from Source Code

If you want to modify or develop the tool further, you can build it from source.

Step 1: Download the Project

1. Clone or download the entire repository to your local machine.
2. After downloading, you will get a .zip file.
3. Right-click the zip file → Select “Properties”.
4. In the Properties panel, check “Unblock”, then click Apply.

⚠️ Why is this necessary?
Windows may block files downloaded from the internet for security reasons.
If you do not unblock the zip file before extracting it, the project may fail to run properly.

5. Extract the zip file.

---

Step 2: Install Visual Studio

1. Download and install Visual Studio.

2. During installation, make sure to select:

   ✅ .NET desktop development

---

Step 3: Open the Project

1. Open the extracted project folder.
2. Locate the file:

YOLIC.sln

3. Open it using Visual Studio.

---

Step 4: Restore NuGet Packages

1. Right-click the Solution in Visual Studio.
2. Click Restore NuGet Packages.
3. Wait for the installation to complete.

After restoration:

• Click the Start button in Visual Studio to build and run the project.

---

🔧 If You Encounter NuGet Errors

If NuGet packages fail to install:

1. In Visual Studio’s top menu, click: Tools → Manage NuGet Packages for Solution
2. Update the required packages to the latest compatible version.
3. Restore packages again and rebuild the solution.

---

📦 Getting Started with Your Own Data

1️⃣ Install the Tool

Install via MSI or build from source as described above.

---

2️⃣ Design Your Own Cell Configuration

You can design your own Interested Cells (COIs) directly within the YOLIC Labeling Tool using the Cell Designer Panel. This feature has been integrated into the software to simplify the configuration process.

Alternatively, you can still manually edit the JSON configuration file if preferred.

Supported COI Types

• Rectangle
• Requires top-left and bottom-right coordinates
• Polygon
• Requires ordered coordinates of all vertices

Example Configuration

{
"Labels": {
 "LabelList": [
   "Bump", "Column", "Dent", "Fence",
   "Creature", "Vehicle", "Wall", "Weed", "ZebraCrossing",
   "TrafficCone", "TrafficSign"
 ],
 "LabelAbbreviation": [
   "Bp", "Cn", "Dt", "Fe", "Ce",
   "Ve", "Wl", "Wd", "ZC", "TC", "TS"
 ],
 "LabelNumber": 11
 },
 "COIs": {
   "COINumber": 3,
   "1": ["rectangle", 0.33962, 0.34583, 0.04009, 0.07083],
   "2": ["rectangle", 0.37971, 0.34583, 0.04009, 0.07083],
   "3": ["rectangle", 0.71226, 0, 0.07075, 0.125]
 }
}

---

3️⃣ Annotate Your Data

Use the YOLIC Labeling Tool to:

• Load your images
• Import your configuration
• Label objects within defined COIs

---

4️⃣ Train a YOLIC Model

Refer to the training repository:

👉 https://github.com/kai3316/YOLIC_code

---

🧠 Learn More About YOLIC

For detailed methodology and experimental results, please refer to the paper:

YOLIC: An Efficient Method for Object Localization and Classification on Edge Devices https://arxiv.org/abs/2307.06689

---

📌 Annotation Example

Cell Configuration for a Road Risk Detection Task

Annotation Tool with Configuration3.json

---

🤝 Contributions & Feedback

We are continuously improving the YOLIC Labeling Tool and welcome:

• Bug reports
• Feature requests
• Pull requests
• Suggestions for improvements

Thank you for your interest in YOLIC! We hope this tool helps you build efficient and lightweight detection systems 🚀