# Infinigen Indoors: Photorealistic Indoor Scenes using Procedural Generation

**Authors:** Alexander Raistrick, Lingjie Mei, Karhan Kayan, David Yan, Yiming Zuo, Beining Han, Hongyu Wen, Meenal Parakh, Stamatis Alexandropoulos, Lahav Lipson, Zeyu Ma, Jia Deng  
**arXiv:** [2406.11824](https://arxiv.org/abs/2406.11824)  
**Date:** 2024-06-17  
**Summary page:** https://www.summarizepaper.com/en/paper/2406.11824v1/

## Key points

- Infinigen Indoors is a groundbreaking system developed by a team of researchers including Alexander Raistrick, Lingjie Mei, Karhan Kayan, David Yan, Yiming Zuo, Beining Han, Hongyu Wen, Meenal Parakh, Stamatis Alexandropoulos, Lahav Lipson, Zeyu Ma and Jia Deng.
- The key feature of Infinigen Indoors is its domain-specific language that allows users to express constraints on scene composition.
- A sophisticated solver generates scene compositions that maximize the satisfaction of these constraints for visually appealing and structurally sound indoor scenes.
- The export tool enables users to utilize generated 3D objects and scenes for training embodied agents in real-time simulators like Omniverse and Unreal.
- Infinigen Indoors is open-source and accessible to developers and creators through https://infinigen.org.
- It offers a powerful tool for creating realistic indoor scenes efficiently, making it valuable for researchers, developers, and artists in digital content creation.

## Summary

Infinigen Indoors is a groundbreaking  developed by a team of researchers including Alexander Raistrick, Lingjie Mei, Karhan Kayan, David Yan, Yiming Zuo, Beining Han, Hongyu Wen, Meenal Parakh, Stamatis Alexandropoulos, Lahav Lipson, Zeyu Ma and Jia Deng. This innovative system expands upon the existing Infinigen platform to focus on the creation of . The key feature of Infinigen Indoors is its , which includes a domain-specific language that allows users to express various constraints on scene composition. A sophisticated solver then generates scene compositions that maximize the satisfaction of these constraints. This approach ensures that the generated indoor scenes are not only visually appealing but also structurally sound and realistic. One of the standout aspects of Infinigen Indoors is its export tool, which enables users to directly utilize the generated 3D objects and scenes for training embodied agents in real-time simulators such as Omniverse and Unreal. This functionality opens up new possibilities for using procedurally generated indoor scenes in virtual environments for training AI models and conducting simulations. Furthermore, the researchers have made Infinigen Indoors , making it accessible to a wide range of developers and creators. Those interested in exploring this cutting-edge technology can visit https://infinigen.org for access to the codebase and instructional videos. Overall,, offering a powerful tool for creating realistic indoor scenes with ease and efficiency. Its innovative features and open-access nature make it a valuable resource for researchers, developers, and artists looking to push the boundaries of digital content creation.