Skip to content

This is a data labeling app for who don't want to share datasets with third party corporations.

License

Notifications You must be signed in to change notification settings

rfatihors/YoloDataLabeler

 
 

Repository files navigation

YOLO Data Labeler

Designed with precision by Melih Taşkın, the YOLO Data Labeler is more than a tool—it's a craft echoing the artistry and innovation reminiscent of the grandeur of Ottoman engineering and the vibrant bazaars of Istanbul.

🌠 Remarkable Features

  • Effortless Image Selection: Just as travelers navigate the mesmerizing lanes of Cappadocia, effortlessly guide your way through the dataset with our intuitive interface.

  • Instantaneous Label Imports:

    {
       "0": "cat",
       "1": "dog"
    }
    
  • Dynamic Image Resizing: Set your requirements and let the application do its magic, reminiscent of the fluid designs seen in Turkish mosaics.

  • Strategic Dataset Segmentation: Employing the power of scikit-learn's split_train_test, carve out your dataset with precision, ensuring each category is well-represented.

  • Intelligent Design: Should challenges arise, like when a label has fewer than three instances, our tool promptly alerts you, echoing the foresight of great Turkish strategists.

🏰 Sovereignty Over Your Data

  • Offline Operation: All activities are local—echoing the impenetrable fortresses guarding Anatolia's rich history.

  • Absolute Control: Your data remains untouched and unshared, echoing the principles of integrity deeply rooted in Turkish customs.

  • Transparent Mechanics: No secret alleys or under-the-table transactions. Your data, your empire.

  • A Truly Independent Tool: Unlike mainstream platforms that might use your data, with YOLO Data Labeler, you're in the Grand Bazaar, and the choice of trade is entirely yours.

🔧 Getting Started

Windows Connoisseurs

Direct Expedition:

Crafting Your Own Path:

```bash
git clone <repository-url>
cd <repository-name>
pip install -r requirements.txt
python main.py
```

Unix/Mac Trailblazers

Your unique journey awaits. While pre-built expeditions for Unix/Mac are on the horizon, the spirit of discovery is at the core:

```bash
git clone <repository-url>
cd <repository-name>
pip install -r requirements.txt
python main.py
```

🛠️ Code Quality Assurance

To ensure the highest standards of coding, our project is integrated with SonarCloud. Not only does this exhibit our commitment to producing top-notch software, but it also provides a clear overview of our code's health and maintainability.

Dive deep into our code's quality metrics and analysis on SonarCloud's project overview.

🌍 Join Our Caravan

We value every comment, critique, and contribution. Thinking of monumental changes? Begin a dialogue with an issue.

📜 Scrolls and Writings

Guarded by the GNU General Public License, delve into the LICENSE for the intricacies.

🙌 Gratitude to the Visionaries

  • A nod to the vast expanse of the open-source realm and the trailblazers of the YOLO model.
  • Our journey is brighter thanks to Ahmet Koyuncu, a guiding star in this venture.

About

This is a data labeling app for who don't want to share datasets with third party corporations.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 100.0%