# 3D scene generation

DistScene is a framework for single-image compositional 3D scene generation that jointly generates the environment and multiple independent objects in a shared scene frame, refines each object in a local space, and distills pretrained object-generation priors into scene generation via automatically composed and rendered synthetic scenes; on indoor and outdoor benchmarks it improves scene-level spatial coherence over the evaluated baselines.

DistScene is a framework for single-image compositional 3D scene generation that jointly generates the environment and multiple independent objects in a shared scene frame, refines each object in a local space, and distills pretrained object-generation priors into scene generation via automatically composed and rendered synthetic scenes; on indoor and outdoor benchmarks it improves scene-level spatial coherence over the evaluated baselines.

Mira-Scene is a compositional single-image 3D scene reconstruction framework that pairs a pixel-aligned Canonical Coordinate Map (CCM) with a scene-space Point Cloud Map (PCM) from monocular geometry estimation to form dense, bounded correspondences, recovers object transformations through robust geometric alignment, and jointly generates object geometry and CCMs with a multimodal diffusion transformer; across indoor, outdoor, synthetic, and in-the-wild scenes it improves layout accuracy over strong baselines, achieving relative gains of 39.8% in 3D-IoU and 16.5% in 2D-IoU over SAM3D on BlendSwap while using limited open-source training data.

Mira-Scene is a compositional single-image 3D scene reconstruction framework that pairs a pixel-aligned Canonical Coordinate Map (CCM) with a scene-space Point Cloud Map (PCM) from monocular geometry estimation to form dense, bounded correspondences, recovers object transformations through robust geometric alignment, and jointly generates object geometry and CCMs with a multimodal diffusion transformer; across indoor, outdoor, synthetic, and in-the-wild scenes it improves layout accuracy over strong baselines, achieving relative gains of 39.8% in 3D-IoU and 16.5% in 2D-IoU over SAM3D on BlendSwap while using limited open-source training data.