Stable Point Aware 3d

🚀 SPAR3D: Stable Point Aware 3D

SPAR3D (Stable Point Aware 3D) is a large reconstruction model. It's based on SF3D and can make real - time edits and generate a textured UV - unwrapped 3D mesh asset from a single image in under a second. It combines a fast point cloud diffusion model with regressive mesh predictions for better 3D object generation control.

🚀 Quick Start

For usage instructions, please refer to our GitHub repository

✨ Features

• Fast 3D Generation: Create a 3D model from a single image in under one second.
• UV - unwrapped and Textured: The generated 3D asset is UV - unwrapped and textured.
• Material Parameter Prediction: Predict per - object material parameters (roughness, metallic) for better reflective behaviors during rendering.
• Improved Backside Modelling: Use a fast point diffusion model for improved backside modelling.

📚 Documentation

Model Description

Model Sources

• Repository: https://github.com/Stability-AI/stable-point-aware-3d
• Tech report: https://arxiv.org/pdf/2501.04689
• Video summary: https://youtu.be/mlO3Nc3Nsng
• Project page: https://spar3d.github.io
• arXiv page: https://arxiv.org/abs/2501.04689

Training Dataset

We use renders from the Objaverse dataset, available under the Open Data Commons Attribution License. Our enhanced rendering method, which mimics real - world image distribution, improves the model's generalization ability. We filter objects based on license review and curate a suitable subset for training.

Intended Uses

• Generation of artworks and use in design and other artistic processes.
• Applications in educational or creative tools.
• Research on reconstruction models, including understanding their limitations.

All uses of the model should comply with our [Acceptable Use Policy](https://stability.ai/use - policy).

Out - of - Scope Uses

The model was not trained to generate factual or true representations of people or events. So, using it for such purposes is beyond its capabilities.

📄 License

• Community License: Free for research, non - commercial, and commercial use by organizations and individuals with annual revenue of US $1,000,000 (or local currency equivalent) or less. If your annual revenue exceeds US $1M, any commercial use of this model or its derivative works requires an Enterprise License from Stability AI. You can request an Enterprise License at https://stability.ai/enterprise. Refer to Stability AI's Community License at https://stability.ai/license for more information.

🔧 Safety

We implement safety measures throughout the model development process, from pre - training to deployment. We've added safety mitigations to reduce severe harm risks. However, developers should conduct their own testing and add more mitigations based on specific use cases. For more on our safety approach, visit our Safety page.

Contact

Please report any issues with the model or contact us:

• Safety issues: safety@stability.ai
• Security issues: security@stability.ai
• Privacy issues: privacy@stability.ai
• License and general: https://stability.ai/license
• Enterprise license: https://stability.ai/enterprise

⚠️ Important Note

For individuals or organizations generating annual revenue of US $1,000,000 (or local currency equivalent) or more, regardless of the source of that revenue, you must obtain an enterprise commercial license directly from Stability AI before commercially using SPAR3D, any derivative work of SPAR3D (such as a “fine tune” model), or their outputs. You may submit a request for an Enterprise License at https://stability.ai/enterprise. Please refer to Stability AI's Community License, available at https://stability.ai/license, for more information.