Mira-Scene: Pixel-Aligned Layouts for Generative 3D Scene Reconstruction
cs.CV, cs.GR
Submitted: 2026-09-20
Updated: 2026-09-22
Comments: Project Page: https://sunyangtian.github.io/Mira-Scene-web/
Project page: https://sunyangtian.github.io/Mira-Scene-web
License: http://creativecommons.org/licenses/by-sa/4.0/
The gist: Single-image 3D object generation can now produce high-fidelity assets, yet accurately placing them into a coherent scene layout remains an open challenge.
Terminology
Abstract
Single-image 3D object generation can now produce high-fidelity assets, yet accurately placing them into a coherent scene layout remains an open challenge. A central difficulty lies in how object layout is represented. Holistic methods absorb placement into a scene-level generation process, sacrificing object-level detail. Compositional methods preserve object fidelity by decoupling geometry from layout, but typically parameterize layout as sparse, unbounded pose variables that are difficult to learn and generalize poorly under scarce scene-level supervision. We present Mira-Scene, a compositional 3D scene reconstruction framework that replaces sparse pose regression with dense, bounded correspondence recovery. At its core is the Canonical Coordinate Map (CCM), a pixel-aligned field that maps each visible object pixel to a surface coordinate in the object's bounded canonical space. When paired with a scene-space Point Cloud Map (PCM) from monocular geometry estimation, CCM induces dense canonical-to-scene correspondences from which object transformations are recovered through robust geometric alignment. Because CCM operates in bounded canonical space, it provides a stable prediction target that can be trained from scalable object-level 3D data without requiring scene-level layout annotations. Mira-Scene further introduces a multimodal diffusion transformer that jointly generates object geometry and CCMs, using modality-specific expert streams with shared attention and positional encoding to promote geometry-layout consistency. Experiments on indoor, outdoor, synthetic, and in-the-wild scenes show that Mira-Scene substantially outperforms strong baselines in layout accuracy, achieving relative gains of 39.8% in 3D-IoU and 16.5% in 2D-IoU over SAM3D, using limited open-source training data.
Sources
- SkyReels-V4: Multi-modal Video-Audio Generation, Inpainting and Editing model
- Automated Creation of Digital Cousins for Robust Policy Learning
- Emerging Properties in Unified Multimodal Pretraining
- Boxer: Robust Lifting of Open-World 2D Bounding Boxes to 3D
- ALIVE: Animate Your World with Lifelike Audio-Video Generation
- LTX-2: Efficient Joint Audio-Visual Foundation Model
- Classifier-Free Diffusion Guidance
- LRM: Large Reconstruction Model for Single Image to 3D
- CUPID: Generative 3D Reconstruction via Joint Object and Pose Modeling
- MegaPose: 6D Pose Estimation of Novel Objects via Render & Compare
- Hunyuan3D 2.5: Towards High-Fidelity 3D Assets Generation with Ultimate Details
- Back to Basics: Let Denoising Generative Models Denoise
- TripoSG: High-Fidelity 3D Shape Synthesis using Large-Scale Rectified Flow Models
- Sparc3D: Sparse Representation and Construction for High-Resolution 3D Shapes Modeling
- Mixture-of-Transformers: A Sparse and Scalable Architecture for Multi-Modal Foundation Models
- PartCrafter: Structured 3D Mesh Generation via Compositional Latent Diffusion Transformers
- I-Scene: 3D Instance Models are Implicit Generalizable Spatial Learners
- Scenethesis: A Language and Vision Agentic Framework for 3D Scene Generation
- Flow Matching for Generative Modeling
- JavisDiT: Joint Audio-Video Diffusion Transformer with Hierarchical Spatio-Temporal Prior Synchronization
Related papers
- Loss Knows Best: Detecting Annotation Errors in Videos via Loss Trajectories
- AnchorWeave: World-Consistent Video Generation with Retrieved Local Spatial Memories
- Benchmarking the Robustness of Foundation Models for Mammography under Domain Shift
- MambaX-Net: Dual-Input Mamba-Enhanced Cross-Attention Network for Longitudinal MRI Segmentation
- TeleOCR: Navigating Document Parsing Across Digital and Camera-Captured Documents
- A Survey on Efficient Vision-Language-Action Models