Reconstruction by Generation: 3D Multi-Object Scene Reconstruction from Sparse Observations

• URL: http://arxiv.org/abs/2604.27106v1
• Date: Wed, 29 Apr 2026 18:51:21 GMT
• Title: Reconstruction by Generation: 3D Multi-Object Scene Reconstruction from Sparse Observations
• Abstract summary: RecGen is a generative framework for probabilistic joint estimation of object and part shapes.<n>It generalizes across diverse object types and real-world environments.<n> RecGen achieves state-of-the-art performance on complex, heavily occluded datasets.
• Score: 29.241988496403383
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Accurately reconstructing complex full multi-object scenes from sparse observations remains a core challenge in computer vision and a key step toward scalable and reliable simulation for robotics. In this work, we introduce RecGen, a generative framework for probabilistic joint estimation of object and part shapes, as well as their pose under occlusion and partial visibility from one or multiple RGB-D images. By leveraging compositional synthetic scene generation and strong 3D shape priors, RecGen generalizes across diverse object types and real-world environments. RecGen achieves state-of-the-art performance on complex, heavily occluded datasets, robustly handling severe occlusions, symmetric objects, object parts, and intricate geometry and texture. Despite using nearly 80% fewer training meshes than the previous state of the art SAM3D, RecGen outperforms it by 30.1% in geometric shape quality, 9.1% in texture reconstruction, and 33.9% in pose estimation.

Related papers

• GARDEN: Gravity-Aligned Reconstruction of Disentangled ENvironments from RGB images [61.2170105836525]
  We propose an RGB-only framework that reformulates reconstruction as physically-grounded scene factorization and outputs a structured hybrid scene representation.<n>The resulting representation combines explicit rigid bodies with a decoupled background, enabling direct physics simulation while preserving visual realism.
  arXiv  Detail & Related papers  (2026-06-02T17:13:01Z)
• SAM3D-Phys: Towards Multi-Object Interactive Simulation in Real World [42.46212677643043]
  SAM3D-Phys is a framework that integrates scene reconstruction with generative 3D priors of SAM3D to recover physically simulatable objects.<n>By recovering complete object geometry and restoring its pose and appearance within the scene, SAM3D-Phys produces clean object representations suitable for physics-based simulation.
  arXiv  Detail & Related papers  (2026-05-28T17:07:51Z)
• 3DReflecNet: A Large-Scale Dataset for 3D Reconstruction of Reflective, Transparent, and Low-Texture Objects [41.25903201175962]
  3DReflecNet is a large-scale hybrid dataset exceeding 22 TB.<n>This dataset spans diverse materials, complex lighting conditions, and a wide range of geometric forms.<n>We design benchmarks for five core tasks: image matching, structure-from-motion, novel view synthesis, removal, and relighting.
  arXiv  Detail & Related papers  (2026-05-11T08:51:28Z)
• ArtLLM: Generating Articulated Assets via 3D LLM [19.814132638278547]
  ArtLLM is a novel framework for generating high-quality articulated assets directly from complete 3D meshes.<n>At its core is a 3D multimodal large language model trained on a large-scale articulation dataset.<n> Experiments show that ArtLLM significantly outperforms state-of-the-art methods in both part layout accuracy and joint prediction.
  arXiv  Detail & Related papers  (2026-03-01T15:07:46Z)
• DreamGrasp: Zero-Shot 3D Multi-Object Reconstruction from Partial-View Images for Robotic Manipulation [12.812918371506209]
  We introduce DreamGrasp, a framework that infers the unobserved parts of a scene.<n>By combining coarse 3D reconstruction, instance segmentation via contrastive learning, and text-guided instance-wise refinement, DreamGrasp circumvents limitations of prior methods.<n>Our experiments show that DreamGrasp not only recovers accurate object geometry but also supports downstream tasks like sequential decluttering and target retrieval with high success rates.
  arXiv  Detail & Related papers  (2025-07-08T03:12:49Z)
• GTR: Gaussian Splatting Tracking and Reconstruction of Unknown Objects Based on Appearance and Geometric Complexity [49.31257173003408]
  We present a novel method for 6-DoF object tracking and high-quality 3D reconstruction from monocular RGBD video.<n>Our approach demonstrates strong capabilities in recovering high-fidelity object meshes, setting a new standard for single-sensor 3D reconstruction in open-world environments.
  arXiv  Detail & Related papers  (2025-05-17T08:46:29Z)
• Gaussian Object Carver: Object-Compositional Gaussian Splatting with surfaces completion [16.379647695019308]
  3D scene reconstruction is a foundational problem in computer vision.<n>We introduce the Gaussian Object Carver (GOC), a novel, efficient, and scalable framework for object-compositional 3D scene reconstruction.<n>GOC leverage 3D Gaussian Splatting (GS), enriched with monocular geometry priors and multi-view geometry regularization, to achieve high-quality and flexible reconstruction.
  arXiv  Detail & Related papers  (2024-12-03T01:34:39Z)
• OpenMaterial: A Large-scale Dataset of Complex Materials for 3D Reconstruction [55.052637670485716]
  We introduce OpenMaterial, a large-scale semi-synthetic dataset for material-aware 3D reconstruction.<n>It comprises 1,001 objects spanning 295 distinct materials, including conductors, dielectrics, plastics, and their roughened variants, captured under 714 diverse lighting conditions.<n>It provides multi-view images, 3D shape models, camera poses, depth maps, and object masks, establishing the first extensive benchmark for evaluating 3D reconstruction on challenging materials.
  arXiv  Detail & Related papers  (2024-06-13T07:46:17Z)
• Total-Decom: Decomposed 3D Scene Reconstruction with Minimal Interaction [51.3632308129838]
  We present Total-Decom, a novel method for decomposed 3D reconstruction with minimal human interaction. Our approach seamlessly integrates the Segment Anything Model (SAM) with hybrid implicit-explicit neural surface representations and a mesh-based region-growing technique for accurate 3D object decomposition. We extensively evaluate our method on benchmark datasets and demonstrate its potential for downstream applications, such as animation and scene editing.
  arXiv  Detail & Related papers  (2024-03-28T11:12:33Z)
• Zero-Shot Multi-Object Scene Completion [59.325611678171974]
  We present a 3D scene completion method that recovers the complete geometry of multiple unseen objects in complex scenes from a single RGB-D image. Our method outperforms the current state-of-the-art on both synthetic and real-world datasets.
  arXiv  Detail & Related papers  (2024-03-21T17:59:59Z)
• Learning Unsupervised Hierarchical Part Decomposition of 3D Objects from a Single RGB Image [102.44347847154867]
  We propose a novel formulation that allows to jointly recover the geometry of a 3D object as a set of primitives. Our model recovers the higher level structural decomposition of various objects in the form of a binary tree of primitives. Our experiments on the ShapeNet and D-FAUST datasets demonstrate that considering the organization of parts indeed facilitates reasoning about 3D geometry.
  arXiv  Detail & Related papers  (2020-04-02T17:58:05Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.