InfiniDream: Open-Ended Photorealistic 3D Worlds Through Procedural Generation
We introduce InfiniDream, a VLM-driven framework for open-ended procedural 3D world generation.
ProofPaper ↗
Key points
- Procedural generation plays a key role in synthetic data creation and game development, offering greater controllability, editability, and diversity than static mesh representations.
- However, VLM-based code generation often produces relatively simple 3D assets, while sophisticated photorealistic procedural systems require extensive manual development.
- First, we propose Symbolically Constrained Asset Generation, which leverages existing procedural algorithms to create novel, photorealistic assets while preserving editability and seed-based diversity.
- Second, we develop an end-to-end text-to-world pipeline that combines semantic scene planning with Deterministic Frustum-Guided Grid Placement to construct large-scale 3D worlds with precise object placement, procedural population, and animation.
Sources (2)
- [1]InfiniDream: Open-Ended Photorealistic 3D Worlds Through Procedural GenerationHugging Face Daily Papers · Oct 19, 12:00 AM
We introduce InfiniDream, a VLM-driven framework for open-ended procedural 3D world generation.
Procedural generation plays a key role in synthetic data creation and game development, offering greater controllability, editability, and diversity than static mesh representations.
- [2]InfiniDream: Open-Ended Photorealistic 3D Worlds Through Procedural GenerationarXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 19, 05:59 PM · same content
Extractive summary: sentences quoted from the sources.
Before this
- Oct 18, 2026FracTS: Hierarchical and Autoregressive Time Series Generation
- Oct 18, 2026VLA-Rolling: Fewer Denoising Steps, Better Control via Adaptive Action Reuse
- Oct 18, 2026Painting Change Map as guidance for generating painting processes
- Oct 16, 2026Welcome Beam: a 501B open model you can own, customize, and scale
- Oct 16, 2026MOSS-Image Technical Report
- Oct 16, 2026Persistent Planar Memory for Video World Models