Summary
LaviGen is a new framework that repurposes 3D generative models for 3D scene layout generation, presented in an arXiv paper by Haoran Feng, Yifan Niu, Zehuan Huang, Yang-Tian Sun, Chunchao Guo, Yuxin Peng, and Lu Sheng. Unlike prior methods that infer object layouts from textual descriptions, LaviGen operates directly in native 3D space and formulates layout generation as an autoregressive process that explicitly models geometric relations and physical constraints among objects, producing coherent and physically plausible 3D scenes. The framework includes an adapted 3D diffusion model that integrates scene, object, and instruction information, together with a dual-guidance self-rollout distillation mechanism that improves efficiency and spatial accuracy. On the LayoutVLM benchmark, LaviGen outperforms state-of-the-art methods, achieving 19% higher physical plausibility and a 65% speedup in computation. Code is publicly available at github.com/fenghora/LaviGen. This post summarizes the paper for the zhichai.net community.
Paper Overview
Field: Computer Vision
Authors: Haoran Feng, Yifan Niu, Zehuan Huang, Yang-Tian Sun, Chunchao Guo, Yuxin Peng, Lu Sheng
Published: 2026-04-17
arXiv: 2604.16299
Abstract
We introduce LaviGen, a framework that repurposes 3D generative models for 3D layout generation. Unlike previous methods that infer object layouts from textual descriptions, LaviGen operates directly in the native 3D space, formulating layout generation as an autoregressive process that explicitly models geometric relations and physical constraints among objects, producing coherent and physically plausible 3D scenes.
To further enhance this process, the authors propose an adapted 3D diffusion model that integrates scene, object, and instruction information, and employ a dual-guidance self-rollout distillation mechanism to improve efficiency and spatial accuracy.
Extensive experiments on the LayoutVLM benchmark show LaviGen achieves superior 3D layout generation performance, with 19% higher physical plausibility than state-of-the-art methods and a 65% speedup in computation.
Links
- Paper: https://arxiv.org/abs/2604.16299
- Code: https://github.com/fenghora/LaviGen
---
*Auto-collected on 2026-04-21.*
This page is an English static mirror generated for search and AI citation.
It may be a full translation or structured summary of the Chinese original.
Canonical interactive discussion lives on the Chinese page:
https://zhichai.net/topic/177618600