CasLayout: Cascaded 3D Layout Diffusion for Indoor Scene Synthesis with Implicit Relation Modeling
Open Access
- 3 July 2026
- journal article
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Graphics
- Vol. 45 (4) , 1-16
Abstract
Synthesizing realistic 3D indoor scenes remains challenging due to data scarcity and the difficulty of simultaneously enforcing global architectural constraints and local semantic consistency. Existing approaches often overlook structural boundaries or rely on fully connected relation graphs that introduce redundant generation errors. Inspired by human design cognition, we present CasLayout, a cascaded diffusion framework that decomposes the joint scene generation task into four conditional sub-stages with explicit physical and semantic roles: (1) predicting furniture quantity and categories, (2) refining object sizes and feature embeddings, (3) modeling spatial relationships in a latent space, and (4) generating Oriented Bounding Boxes (OBBs). This decoupled architecture reduces data requirements and enables flexible integration of Large Language Models (LLMs) and Vision Language Models (VLMs) for zero-shot tasks such as image-to-scene generation. To maintain physical validity within complex floor plans, we explicitly model building elements ( e.g. , walls, doors, and windows) as conditional constraints. Furthermore, to address the high entropy of dense relation graphs, we introduce a sparse relation graph formulation aligned with human spatial descriptions. By encoding these sparse graphs into a compact latent space using a bidirectional Variational Autoencoder (VAE), the proposed framework provides enhanced relational controllability, allowing generated layouts to better respect functional organization. Experiments demonstrate that CasLayout achieves state-of-the-art performance in fidelity and diversity while enabling improved controllability in practical applications.Keywords
All Related Versions
This publication has 37 references indexed in Scilit:
- SceneFlow: Synthesizing indoor scenes via geometry-enhanced flow matchingComputer Aided Geometric Design, 2025
- I-Design: Personalized LLM Interior Designer2025
- FOREST2SEQ: Revitalizing Order Prior for Sequential Indoor Scene Synthesis2024
- AnyHome: Open-Vocabulary Generation of Structured and Textured 3D Homes2024
- PhyScene: Physically Interactable 3D Scene Synthesis for Embodied AI2024
- Holodeck: Language Guided Generation of 3D Embodied AI Environments2024
- LEGO-Net: Learning Regular Rearrangements of Objects in Rooms2023
- SceneFormer: Indoor Scene Generation with Transformers2021
- Fast and Flexible Indoor Scene Synthesis via Deep Convolutional Generative Models2019
- ScanNet: Richly-Annotated 3D Reconstructions of Indoor Scenes2017