PulseAugur
EN
LIVE 12:44:12

HomeWorld framework generates controllable, interactive whole-home scenes

By PulseAugur Editorial · [2 sources] ·

Researchers have developed HomeWorld, a novel framework for generating realistic and interactive whole-home scenes. This system utilizes a large language model trained on 300,000 floorplans to create diverse layouts, followed by image generation models for furniture placement. A VLM-based refiner and 3D generative model further enhance object placement and asset replacement, preparing the scenes for embodied AI simulations and interior design applications. The project will also release a floorplan dataset and 5,000 furnished scenes. AI

IMPACT Enables more realistic simulations for embodied AI and aids in interior design.

RANK_REASON Academic paper detailing a new framework for scene generation.

Read on arXiv cs.CV →

AI-generated summary · Google Gemini · from 2 sources. How we write summaries →

COVERAGE [2]

  1. arXiv cs.AI TIER_1 English(EN) · Hongsheng Li ·

    HomeWorld: A Unified Floorplan-to-Furnished Framework for Generating Controllable, Densely Interactive Whole-Home Scenes

    Indoor scene generation is crucial for robot simulation and modern interior design. However, complex layouts together with scarce 3D scene data make learning-based generation challenging. Existing methods often rely on hand-crafted rules or focus on isolated sub-tasks (e.g., floo…

  2. arXiv cs.CV TIER_1 English(EN) · Wenbo Li, Xiaoliang Ju, Zipeng Qin, Rongyao Fang, Hongsheng Li ·

    HomeWorld: A Unified Floorplan-to-Furnished Framework for Generating Controllable, Densely Interactive Whole-Home Scenes

    arXiv:2606.06390v1 Announce Type: new Abstract: Indoor scene generation is crucial for robot simulation and modern interior design. However, complex layouts together with scarce 3D scene data make learning-based generation challenging. Existing methods often rely on hand-crafted …

RELATED ENTITIES

RELATED TOPICS