Overview

3D reconstruction studies how to infer the shape and appearance of objects or scenes, along with camera positions and motion, from partial observations. Methods range from structure-from-motion and multi-view stereo to neural radiance fields, Gaussian splatting, signed distance fields, meshes, and feed-forward models that predict geometry directly; some also use diffusion priors, active view selection, or physical sensor models. Applications include mapping and robotics, digital twins, medical imaging, and human or object modeling, with evaluations often comparing both rendered views and geometric accuracy on indoor, outdoor, and specialized datasets. A central challenge is that realistic novel-view images do not always imply accurate, metrically consistent geometry, especially with sparse or unposed views, changing scenes, difficult materials, or limited computation.

Latest papers

1,033 papers76 papers in the last four weeks, down 37% on the four weeks before. 0.6% of all new papers.

Nov 3, 2026

Sparse-View 3D Coronary Reconstruction from Vessel Segmentations Using a Diffusion Prior

Reconstructing three-dimensional (3D) coronary anatomy from sparse angiographic views could support percutaneous coronary intervention, but limited angular coverage and imperfect vessel segmentations introduce substantial geometric ambiguity. We propose a vessel-specific reconstruction framework combining a learned 3D diffusion prior, visual-hull initialization, and projection-based data consistency. The unconditional diffusion model is trained on coronary vessel masks without paired 2D--3D data and reused across different view counts without retraining. During inference, a scaled visual hull initializes reconstruction, while projection-based updates constrain diffusion refinement using patient-specific vessel observations. We evaluated the framework on simulated angiograms derived from ImageCAS computed tomography angiography volumes using two, four, and eight views at two contrast levels. Across 50 evaluation coronary-tree cases, the proposed method achieved higher mean Dice and clDice than visual hull, deep image prior (DIP), and hash-grid implicit neural representation (INR) at all view counts under reduced contrast. With eight views, improvements over DIP, the best-performing baseline at lower contrast, were approximately 0.13 in Dice and 0.22 in clDice. At higher contrast, performance remained competitive with hash-grid INR, the best-performing eight-view baseline. The preferred diffusion starting timestep depended on observation quality and view count, with excessive refinement reducing accuracy at higher contrast. Projection guidance improved both metrics across all conditions. Averaging stochastic reconstructions further improved accuracy, while sample disagreement localized errors with average-precision gains of 0.14--0.21 over error prevalence. These findings support diffusion-prior reconstruction from sparse vessel observations, although clinical validation remains necessary.
Nov 3, 2026

Predicting Void Points in Self-Occluding Forest Volumes

Synthetic Aperture Imaging (SAI) enables drones to observe forest canopies by warping and averaging multiple displaced views at a common depth. However, occlusion is suppressed rather than eliminated, causing poorly observed regions to appear occupied and leading to systematic biomass overestimation. The Deep Forest project cannot distinguish reconstruction artifacts from real vegetation because its training data relies on measured reflectance points, excluding empty space and preventing the network from learning explicit occupancy boundaries. This thesis addresses the problem by modifying training labels rather than the imaging process. Each simulated scene is rendered twice: as a 9x9 grid of top-down captures for focal-stack generation and as 550 depth-clipped slabs derived from scene geometry to provide occupancy labels, explicitly representing void space as a negative class. A Gazebo pipeline generates procedural forest stands containing 80-200 trees across three species. A pretrained Pix2Pix model translates rendered RGB images into a synthetic near-infrared band, and Light Field Rendering produces aligned focal stacks of 550 planes across three channels. Four architectures were evaluated on a held-out scene. The plain 2D U-Net achieved an IoU of 0.6290, outperforming the volumetric 3D U-Net (0.5707). Established reconstruction methods performed substantially worse: COLMAP achieved 0.3446 and DepthAnything3 0.2959, both after their respective optimal dilation sweeps. An RGB-input variant of the volumetric architecture achieved the highest IoU of 0.6514 without dilation or per-scene calibration. These results highlight the limitations of conventional reconstruction methods, which recover visible surfaces but struggle to infer occupancy within closed canopies.
Nov 3, 2026

Mamba3R: Fixed-Capacity Selective State Memory for Long-Sequence 3D Reconstruction

Streaming 3D reconstruction must maintain geometric quality over long inputs within a fixed GPU budget. A pretrained recurrent model provides bounded state, but freezing its backbone also fixes its learned state-observation interactions. We introduce Mamba3R, a jointly designed state-memory adapter that provides a trainable adaptation path without updating the backbone or task heads. A selective state-space mixer jointly transforms native state features and a fixed-capacity auxiliary memory, explicit read and write paths carry information across frames, and a decomposed feedback path regulates what the added memory feeds back. Only 1.28 million adapter parameters are trained. Under a shared periodic-reset protocol, it lowers trajectory error and improves accumulated surface and depth accuracy over recurrent baselines on public indoor and outdoor datasets and robot-recorded traversals, gaining up to 40 percent on one metric. Matched comparisons support both the choice of the mixing operator and the utility of carrying the auxiliary memory across frames. Allocated memory remains unchanged and throughput moves by at most 1.4 percent from one to two thousand frames. Replaying its recorded traversals offline, the platform's RTX 4060 averages 11.7 frames per second within 5.1 GiB with the same checkpoint, a mean throughput covering the 10 Hz sensor rate.
Nov 2, 2026

3D dark matter map of the JWST COSMOS-Web field using diffusion-based generative priors

Mapping the three-dimensional distribution of dark matter is key to tracing the growth of cosmic structure. Two-dimensional weak lensing mass maps recover its projected distribution, but they blend structures at different distances along the line of sight. We reconstruct the three-dimensional matter distribution in the 0.54 deg2^2 COSMOS-Web field using deep James Webb Space Telescope imaging, with an effective source density of 121 galaxies arcmin−2^{-2} We combine a physical weak lensing forward model with a diffusion-based Bayesian generative prior learned from cosmological simulations, yielding a posterior distribution of the matter field across 20 lens planes to z≃2z \simeq 2, doubling the redshift reach of previous Hubble Space Telescope-based 3D reconstructions and approaching cosmic noon, an epoch of rapid galaxy growth. The reconstruction separates structures that overlap in projection and recovers non-Gaussian matter structure across a range of spatial scales, including over-densities associated with independently identified galaxy clusters that were not detected in earlier weak lensing maps. Posterior sampling quantifies the range of three-dimensional matter distributions allowed by the data, enabling a probabilistic visualization of the cosmic web while propagating uncertainties in its inferred structure. Null and simulation-based tests show that the recovered structures are supported by the lensing signal and are consistent with expectations from the fiducial ΛΛCDM cosmology. These results extend three-dimensional weak lensing mapping into the high-redshift, non-Gaussian regime and establish a route towards probabilistic mapping of the evolving cosmic web with current wide-field lensing surveys.
Nov 2, 2026

Visual-to-Executable Procedural Reconstruction of Buildings

Procedural modeling offers compact, editable, and reusable representations for architectural content, but building such models still depends heavily on manually authored grammars. We study image-based procedural building reconstruction from in-the-wild images, with the goal of recovering an executable procedural description that can be instantiated as an editable 3D building asset. Because a single image provides only partial and view-dependent evidence, we focus on recovering the dominant visible facade(s) and attach the recovered facade program to a simple mass model for building-level instantiation. Our method decomposes this task into two structured prediction stages. First, a Global Structure Layout Tree recovers facade-scale split-and-repeat organization through a fine-tuned layout model with render-guided visual-symbolic reflection. Second, a Component Parameterization schema maps representative facade tiles to structured procedural attributes for local architectural component synthesis. These two representations are then deterministically compiled into executable Computer Generated Architecture (CGA) programs, avoiding the brittleness of direct low-level program generation. Experiments show that the proposed framework handles both regular facades and challenging in-the-wild images, while maintaining high structural consistency, procedural executability, and fine-grained architectural detail. The recovered programs further enable controllable and reusable 3D building asset generation.
Nov 2, 2026

3D Gaussian Splatting within Self-Learned Neural View-Dependent Color Fields

3D Gaussian Splatting (3DGS) has become a vital tool in novel view synthesis. 3D Gaussian functions with attributes like color, opacity, and scale are stacked together to explicitly model a radiance field. However, discrete 3D Gaussians usually produce poor depth with few geometry details but lots of floaters, which hinders the effective modeling of smooth radiance fields. To address this issue, we propose to learn a radiance field within a continuous view-dependent color field parameterized by neural networks. With continuous constraints in color, the bias on low-frequency signals of neural networks can highly encourage the geometry modeling to contribute to high-frequency color variations on images, getting rid of solely relying on the color attribute itself. Moreover, we impose a constraint to improve the Lipschitz continuity of the neural color field, making large changes of each Gaussian's color also match with large changes of its position in terms of a ratio. To this end, we additionally introduce novel self-learning strategies to improve the learning of the neural view-dependent color field without any external knowledge or priors, aiming for better generalization on unseen scenes. Our numerical and visual comparisons on widely used benchmarks justify our idea and show better ability of high fidelity geometry recovery in 3DGS. Project page: https://github.com/yangyixi21/Neural-Color-Fields-3DGS.
Nov 2, 2026

GPU-Native Real-Time Textured Mesh Reconstruction with Seam-Aware Multi-View Radiometric Correction

We present a GPU-native volumetric-video system for real-time reconstruction of graphics-compatible textured triangle meshes from multiple RGB-D sensors. Existing TSDF systems accelerate fusion and mesh extraction, but extending them to per-frame textured meshing adds CPU mesh processing, CPU-side texturing work, or extra CPU-GPU transfers between stages. After input staging, our method keeps the reconstruction and texturing stages (TSDF fusion, Marching Cubes, connected-component noise removal, view-based texture mapping, inpainting, and radiometric seam correction) resident on the GPU; no intermediate mesh or image data returns to the host between these stages. A fixed-size-array representation of mesh topology gives constant-time neighbor access on the GPU and makes the downstream mesh-level operations data-parallel. Because we texture each face from a single camera, exposure seams that vary smoothly across each image remain; we estimate a per-camera low-frequency gain field and bake it into the atlas on the GPU. On the CMU Panoptic dataset, our method is faster than the baselines at every tested voxel resolution and number of sensors, achieves peak speedups of more than an order of magnitude over CUDA-accelerated nvblox and CPU-based Open3D, and approaches an offline textured-meshing baseline in perceptual re-rendering quality.
Nov 2, 2026

Solv3R: 3D Looped Transformer as a Multi-view Solver

Classical structure-from-motion reconstruction proceeds iteratively, refining camera poses and scene geometry from a current estimate. This allows maps to easily expand without recomputing already computed views, or geometry estimates to improve as more observations are provided. Most geometric foundation models, however, predict entire geometry in a single forward pass or update a recurrent state as frames arrive sequentially, making this expansion and refinement non-trivial. We introduce Solv3R, a looped transformer that learns to continue solving multi-view geometry from its current state. Importantly, its update operator is independent of the iteration count, allowing previously refined and newly introduced views to be jointly updated regardless of how ``solved'' they are. We train the operator to preserve accurate reconstructions under continued iteration and correct inaccurate states, enabling refinement of fixed image collections and integration of new observations. Representations of solved views can also be cached once and reused as geometric context across iterations, enabling compact and efficient key-value caching. We analyze the solver dynamics and compare it with the state-of-the-art, demonstrating that a single learned solver supports offline reconstruction, incremental reconstruction, relocalization, and streaming without task-specific training or modules. The code and models will be made publicly available at https://ubc-vision.github.io/solv3r/.
Nov 1, 2026

Learning from high-resolution 2D images to reconstruct 3D microstructures

High-resolution three-dimensional (3D) microstructure characterization remains difficult to undertake at scale because volumetric imaging at the required spatial resolution is expensive, time-consuming or destructive. In many settings, however, fine-resolution two-dimensional (2D) imaging and lower-resolution volumetric measurements are both readily available. We introduce a Privileged Cross-modal Relational Distillation (PCRD) framework that leverages fine-resolution 2D observations during training to improve specimen-specific 3D reconstruction from lower-resolution volumetric imaging alone. We demonstrate PCRD on pultruded carbon fiber-reinforced polymer composites, using scanning electron microscopy as the privileged modality and micro-CT as the inference input. The framework transfers spatial organization by matching similarity distributions defined around corresponding spatial anchors in the teacher and student representations, avoiding direct alignment of heterogeneous feature vectors. In 2D cross-sectional reconstruction, PCRD improves mAP by 1.66 percentage points over direct micro-CT training. Independent 3D validation against high-resolution synchrotron X-ray CT data shows close agreement in fiber volume fraction, diameter and local spacing. The PCRD reconstruction matches 89.97% of the reference fiber centerline length, compared with 75.08% for the interpolated low-resolution volume.
Nov 1, 2026

EgoMo3R: Joint Egocentric Motion and Scene Reconstruction

Understanding holistic 3D human motion and the surrounding environment from egocentric video is fundamental for applications in AR/VR and robotics. Although these tasks are inherently coupled, existing methods treat them separately. In particular, egocentric motion reconstruction methods typically assume access to metric, gravity-aligned camera trajectories from specialized inertial sensors. Recovering such trajectories from monocular egocentric video remains a significant challenge due to rapid head movements and motion blur. We present EgoMo3R, a joint optimization framework that simultaneously reconstructs 3D human motion and scene geometry from casual egocentric video. It leverages the robustness of low-level visuo-geometric primitives, including monocular depth, optical flow, and point tracks, integrated with a diffusion-based human motion prior via first-order optimization. We propose an alternating optimization scheme that establishes a synergistic feedback loop: scene reconstruction provides physical grounding and trajectory conditioning for the motion prior, while the motion prior regularizes the camera trajectory and anchors the scene in a metric, gravity-aligned world frame. We evaluate EgoMo3R on the highly dynamic sequences of the EgoExo4D and Nymeria datasets and show that it outperforms specialized baselines in both 3D scene reconstruction and egocentric motion estimation.
Oct 31, 2026

A Reproducible Reconstruction and Evaluation Framework for Photorealistic Lunar Analog Scene Reconstruction Toward Digital Twins

In this paper, we present a reconstruction pipeline that empirically quantifies the photorealistic scene reconstruction accuracy from both human-controlled and rover mast camera monocular feeds against physical ground truth as a function of Lunar traversal effects (e.g., motion blur, haze due to lofted regolith, trajectory bias, and sparse views). First, we benchmark six variants of 3D Gaussian Splatting (3DGS) across three Lunar analog scenes on held-out test sets using standard photometric metrics. We then establish how under traversal effects, this test-set-specific Novel View Synthesis (NVS) performance does not proportionally scale to the same photometric metrics in the reconstructed DTs. Our results quantify that splat training metrics alone are insufficient predictors of Cross-Reconstruction Photorealistic NVS Fidelity (CRPNVS), and that factors such as rover trajectory bias and geometric scene coverage play a critical role in determining the photorealistic quality of the reconstructed scenes targeting Lunar analog Digital Twins. Additionally, we pilot study Coverage Score (ScS_c), a geometry-based diagnostic derived from COLMAP poses and point cloud structure that provides a meaningful heuristic about the highest achievable CRPNVS structural fidelity, independently of splat training performance in the absence of a human-controlled reference. These findings establish a standardized, human control-referenced evaluation pipeline for photorealistic scene reconstruction for Lunar-analog Digital Twin development and future planetary exploration. Code and project page: https://rose-lab-admin.github.io/rose-lunar-dt-rec
Oct 31, 2026

3D Neural Rendering for Automated Ultrasonic Laser Scans

Laser doppler vibrometer (LDV) scans can be used to detect defects and cracks in manufactured parts before they are deployed. This work presents a method for reconstructing the steady-state surface wavefield of a three-dimensional part by combining automated robotic manipulation with neural rendering. A six-degree-of-freedom (6-DOF) robotic manipulator (UFactory xArm6) with an embedded piezoelectric transducer presented an acoustically excited test article to a stationary LDV at a sequence of known orientations, yielding a set of complex-valued 2D scans with poses derived analytically from the arm kinematics rather than from image-based pose estimation. These scans were used to train a novel complex-nerfacto model, a modified Neural Radiance Field (NeRF) pipeline that replaces the standard three-channel RGB output head with a two-channel output. Complex-nerfacto represents the real and imaginary velocity components of a wavefield, preserving the complex measurement without quantization. Reconstructions of a welded aluminum cylinder and a welded T-bracket were evaluated against measured scans using Z-normalized cross-correlation (ZNCC). The complex-valued pipeline achieved a mean ZNCC of 0.60 at held-out viewing angles, compared with 0.28 for a baseline RGB nerfacto model trained on rendered images of the same data. The resulting surface wavefield reconstructions supported defect detection and characterization on complex components. We identify geometry reconstruction from sparse LDV scans as the remaining obstacle to practical deployment.
Oct 27, 2026

Efficacy of Monocular Depth Estimation in Lunar Environments

Recent advances in monocular depth estimation have demonstrated highly accurate 3D reconstructions within in-distribution terrestrial domains. This study investigates the efficacy of several state-of-the-art approaches to reconstruct out-of-distribution lunar environments using the NASA Ames Polar Optical Lunar Analog Reconstruction (POLAR) stereo dataset and Apollo 17 surface imagery. By leveraging these datasets, we establish a robust analog-to-field validation pipeline that evaluates quantitative performance against ground-truth physical analogs and qualitative performance against historical field data. Our results indicate that select state-of-the-art techniques can successfully generalize to produce high-fidelity relative 3D reconstructions of the unstructured lunar surface.
Oct 27, 2026

GOFish: Full Field-of-View Gaussian Splatting and Mesh Extraction

Robots, autonomous vehicles, and consumer 360° cameras use fisheye lenses for wide fields of view (FoVs). Simulation-ready digital twins require photorealistic rendering across the full FoV and an aligned collision mesh. However, prior fisheye Gaussian Splatting methods only rasterize up to 180° and do not produce meshes, while Gaussian surface reconstruction methods assume pinhole training images. We present GOFish, the first 3D Gaussian Splatting surface reconstruction method to train directly on fisheye images, rasterize FoVs beyond 180°, and extract an aligned mesh from the same scene. Calibrated densification recovers the surface across the entire FoV, while Calibrated Mip Filtering and Wide-Angle Appearance Modeling keep pixel blur and lens falloff out of the scene and mesh. Across the 200° FIORD and 205° FullCircle benchmarks, GOFish improves PSNR over the best prior method by 2.28 and 1.97 dB, respectively. It also outperforms fisheye rendering baselines within their own FoVs and achieves the highest mesh F-score on FIORD and the main ScanNet++ set. Together, the Gaussian scene and mesh form a LiDAR-free, simulation-ready digital twin that renders in real time, ray traces the full FoV, and provides collision geometry.
Oct 27, 2026

Honeycomb-Surfels: Compact Hexagonal Primitives for Efficient Scene Rendering and Reconstruction

Novel view synthesis and 3D reconstruction have advanced rapidly with radiance fields and splatting, yet their memory and computational demands limit their use in resource-constrained settings. We introduce Hexels: novel planar hexagonal surface elements with a first-order spatial color field for efficient 3D scene and 2D image representation. Hexels exploit locally planar regions with smoothly varying appearance using fewer, larger primitives. Their hexagonal support provides an area- and boundary-efficient domain, while the local color field captures appearance variation within each primitive. Across 34 scenes, Hexels improve PSNR over memory-matched 2DGS by 4.53-7.38 dB across datasets while requiring lower training VRAM. We further ablate Hexels by showing that per primitive, Hexels outperform 2DGS by upto 4.74 dB while using lower training VRAM and having 2.31 times faster FPS. Even in unconstrained settings, Hexels offers complementary quality-memory trade-offs. For efficient 3D surface reconstruction, Hexels reduce error by 28.9% compared with memory-matched 2DGS across 15 DTU scans. On 6 Tanks and Temples surface scenes, Hexels achieves the highest mean F1 among compared methods (34.38%, versus 31.72% for 2DGS). We further also show that Hexels, as a general primitive, improves 2D image representation over Image-GS. Hexels push the Pareto frontier on the quality-efficiency trade-off of surface-aligned primitives.
Oct 27, 2026

Learning to Code the World: Harness Evolution for Compositional 3D Reconstruction

Reconstructing editable and compositional 3D scenes from a real image requires jointly reasoning about geometry, object structure, and spatial relationships. We introduce SceneEvolve, a framework that represents 3D scenes as executable code and improves reconstruction by evolving the harness around a small open-source agent. SceneEvolve combines SceneAct, a compositional semantic action interface built on parameterized primitives, hierarchical composition, and semantic spatial relations, with SceneHarness-RSI, which leverages a stronger teacher agent to iteratively evolve the harness using feedback from development-set reconstructions. Throughout harness evolution, the model parameters, action semantics, and execution backend remain fixed. The resulting harness provides a reusable reconstruction procedure that enables the small open-source model to reconstruct unseen scenes without parameter updates or teacher assistance. Across two in-domain datasets and one out-of-domain dataset, SceneEvolve with the comparatively compact Qwen3.8-27B model improves all four evaluated reconstruction metrics while reducing mean inference time by 47.0% relative to the base model. More notably, despite relying on a substantially smaller model, SceneEvolve achieves 32.3% higher reconstruction quality than the large proprietary GPT-5.6-Sol model, while reducing mean equivalent deployment cost by 59.8%. Furthermore, the harness evolved using Qwen3.8-27B also generalizes to Kimi-K3, achieving comparable reconstruction performance to GPT-6-Astra. The project page is https://haoyu-agent.github.io/SceneEvolve.
Oct 27, 2026

HoloRecon: Incremental Reconstruction of Long-Range Panoramas with Global Consistency

Long-sequence panoramic reconstruction requires detailed scene geometry and globally consistent camera poses across distant revisits. Panoramic SLAM and incremental SfM rely on feature correspondences, but repeated, similar structures and small partial overlap make reliable long-range connections difficult to establish, leaving accumulated drift unresolved. Feed-forward models offer detailed local geometry, but their high computational costs limit inference window size, necessitating additional constraints for global consistency. This motivates combining continuous feature-based tracking and sparse mapping with rapid, high-quality local reconstruction from feed-forward models to achieve accurate localization and globally consistent dense mapping over long sequences. We therefore propose a panoramic reconstruction framework that couples an ORB-based sparse panoramic mapping frontend with a fine-tuned feed-forward model. Our key insight is that predicted local geometry enables 3D verification of feature matches across distant revisits, yielding long-range camera constraints even between regions. We calibrate predicted depths with sparse landmarks and use verified connections to optimize the global pose graph, updating dense geometry with the corrected camera poses. We also introduce PanoRecon, a seven-scene benchmark with repetitive architecture, large and small loops, and reference camera poses and LiDAR geometry for evaluating global consistency. Our method achieves the lowest mean absolute trajectory error and high F-1 Score among the evaluated methods. Project Page:https://zhuoxiaoli.github.io/HoloRecon_page/
Oct 27, 2026

Ground Sample Distance Constrained k-Coverage Viewpoint Planning via Set Multi-Cover Formulation

Drone-based 3D reconstruction and inspection require viewpoints that satisfy image-quality constraints while observing each surface region multiple times, which overlap-based planning heuristics do not guarantee. We study ground sample distance (GSD)-constrained kk-coverage viewpoint selection for structures represented by surface voxels. Candidate viewpoints are generated along outward surface directions, screened with distance, viewing-angle, field-of-view, and occlusion tests, and augmented for voxels with insufficient visibility. Selection is then posed as set multi-cover with per-voxel coverage demands. We propose a dynamically weighted greedy heuristic that prioritizes remaining coverage deficits and scarce viewing alternatives, and compare it with sequential overlap-based pruning and an integer linear programming (ILP) formulation on the same candidate pools. Across two structures, four voxel resolutions, and three coverage requirements, the proposed approach achieves full or near-full kk-coverage with compact viewpoint sets, using only 3.00--16.39% more viewpoints than ILP while offering an interpretable selection process based on coverage deficits and viewpoint availability.
Oct 26, 2026

Any3R: Feedforward 3D Reconstruction from Any Camera Type

Geometry foundation models recover camera poses and dense geometry from a set of images with a feed-forward network, but they expect pinhole images. That assumption is inside the backbone too: the positional encoding and the training loss both bind an image position to a direction in space, and both depend on the lens. We condition both on the dense ray field, the unit direction through every pixel, which describes any central camera without a lens type, distortion coefficients or a bound on field of view. Angular coordinates give the rotary encoding the measured ray angle of each patch, evaluated per view, and solid-angle weighting scales the point loss by the area each pixel covers on the sphere, so neither the encoding nor the supervision depends on the lens. Any3R is therefore a feed-forward model that takes any central camera, and is robust to mixing many camera types per scene, with no retraining per lens. On a benchmark of thirteen central cameras across three datasets, where every method reconstructs with its own predicted camera poses, it is the most consistently accurate method. Any3R enables scalable 3D reconstruction from heterogeneous crowdsourced and archival imagery collections sourced across diverse camera types. Project page: https://jingong-chen.github.io/Any3R/
Oct 26, 2026

SceneAgent: An Agentic Pipeline from 3D and 4D Captures to Simulation Environments

Visual captures provide detailed scene appearance, but converting them into interactive simulation environments requires object reconstruction, physical modeling, and correction of intermediate errors. We present SceneAgent, an agentic pipeline that coordinates these steps through simulator-based visual inspection and repair. The pipeline processes 3D Gaussian Splatting (3DGS), photogrammetry, LiDAR captures, AI-generated 3DGS scenes, and 4D videos of dynamic scenes, and combines object segmentation, physical parameter estimation, part decomposition, articulation, and deformable simulation. To support policy learning, SceneAgent generates diverse object variants ("digital sisters") and scripted demonstrations. On a cluttered bin scene from GraspClutter6D, SceneAgent raises F1 at 1 cm for every coding agent with scorable outputs both with and without its skills, and its reconstructions are better than SimFoundry: all agent configurations exceed SimFoundry's F1 score, with a best of 60.75% versus 6.9%. On three dynamic scenes from 4DCodeBench, SceneAgent raises the mean dynamic IoU of Gemini 4 Argon from 0.86% to 99.09%. We further demonstrate diverse simulated interactions and preliminary policy transfer to a real robot.
Oct 25, 2026

RIDE-Splat: Reflection Removal for In-the-Wild Driving Environments

Driving video captured through a windshield violates the world-space consistency assumed by 3D Gaussian Splatting (3DGS): the road scene exhibits parallax while cabin reflections remain attached to the camera, causing standard 3DGS to embed reflected content into the reconstructed scene. We introduce RIDE-Splat, a pipeline for reflection-free 3D reconstruction from posed driving video without clean images, reflection masks, or pretrained networks. RIDE-Splat assigns world-consistent transmission through the windshield to a 3DGS scene and the dominant camera-attached reflection to one shared, signed RGB image. It builds the clean reconstruction in two passes. First, it learns the 3D geometry and the signed reflection image, representing both added radiance and transmission attenuation. Then, it uses the reflection-removed training images to fit a clean 3DGS model. At inference, the reflection image is discarded, preserving the standard 3DGS rendering path. We also introduce the RIDE-Splat Dataset, the first 3D reconstruction benchmark for reflection removal from driving video -- with real driving scenes, physical windshield reflections in training views, and separately captured non-reflective evaluation images. Across the RIDE-Splat Dataset, XLD, and Wayve, RIDE-Splat outperforms 3DGS, DC-Gaussian, and RobustSplat on all nine aggregate metrics, improving PSNR by up to 3.59 dB and reducing LPIPS by up to 0.0876. We achieve the highest metrics in 38 out of 42 per-scene comparisons and removes windshield reflections that remain visible in the baselines while preserving the transmitted scene.
Oct 25, 2026

Hybrid Mamba-Transformer Network for Single-Shot 3D Reconstruction from Fringe Patterns

Deep learning facilitates rapid single shot three dimensional (3D) reconstruction using fringe projection profilometry (FPP), but recovering accurate surface height from a single high frequency fringe image remains challenging because both long-range spatial consistency and fine geometric details must be preserved. Convolutional networks are limited in modeling long range dependencies, whereas window-based Transformers establish global interactions only progressively with increasing depth. To address these limitations, a hybrid Mamba Transformer network is proposed for fringe to height regression. The network adopts a Swin based U shaped architecture and introduces visual state space modeling at the intermediate resolution encoder stage. An axial and dilated vision state space (ADVSS) block is further developed to enhance prescan feature encoding through local, dilated, and axial depthwise convolutions while retaining four directional selective scanning. This design combines linear complexity long range modeling with direction aware preservation of local fringe structures. Experiments on a publicly available FPP dataset show that the proposed method outperforms hNet, SwinConvUNet, UHRNet, and HiDNet in mean absolute error(MAE), root mean square error (RMSE), absolute relative error, and structural similarity index measure (SSIM). Using only one deformed fringe image, the proposed method achieves an RMSE of 0.548 mm and an SSIM of 0.9985, reducing RMSE by 65.2% and 59.6% relative to UHRNet and HiDNet, respectively. These results demonstrate the effectiveness of combining selective state space modeling with hierarchical window based attention for accurate single-shot FPP reconstruction.
Oct 24, 2026

FreeView: Robust 3D Reconstruction from Uncurated Views

Reconstructing a coherent 3D scene from uncurated image collections remains challenging due to the presence of irrelevant or non-overlapping views. Recent work has shown that feed-forward reconstruction models contain internal view-consistency signals that support distractor rejection. However, reference-conditioned filtering relies on a curated clean reference, which is not guaranteed in practical image collections. To address this limitation, we introduce FreeView, a training-free framework that jointly selects a filtering reference and rejects distractor views from an uncurated collection. FreeView leverages collection-level consensus to identify a representative view that serves as a reliable reference for filtering. It then adapts view rejection to each input collection by combining complementary visual and geometric evidence. This approach preserves useful views under substantial appearance and viewpoint variations while rejecting distractors. FreeView integrates with both VGGT and pi^3, removing the need for a manually curated reference or manual calibration of view rejection. Extensive experiments across multiple datasets demonstrate high clean-view retention and distractor rejection rates, leading to reliable 3D reconstruction from uncurated views.
Oct 24, 2026

CADonatello: Towards Real-World Multi-Part 3D CAD Reconstruction

Existing methods for CAD reconstruction from 3D scans achieve strong performance on standard benchmarks, yet their effectiveness often degrades in real-world scenarios where objects consist of multiple parts and scan artifacts such as noise. A key reason is that current CAD reconstruction datasets predominantly contain single-part models and clean geometric inputs, limiting the ability of learned models to generalize to realistic settings. In this work, we introduce CADonatello, a part-aware framework for reconstructing parametric CAD models from real-world 3D scans. Our approach first decomposes the input geometry into structurally meaningful components using a foundation-scale 3D part segmentation model, and then reconstructs each component independently through a cross-section-driven pipeline that performs sketch parameterization followed by an extrusion. To improve the robustness of sketch parameterization to scan-derived cross-sections, we further introduce CC3D Cross-Sections, a dataset of paired rasterized cross-section images from realistic scans and their corresponding parametric sketches derived from aligned CAD models. This dataset enables effective domain adaptation from clean CAD sketch images to scan-derived cross-section images. Extensive experiments demonstrate that the proposed approach improves generalization to noisy scans and enables reliable reconstruction of complex multi-part objects across multiple real-world CAD reverse engineering benchmarks.
Oct 22, 2026

OmniRay: Camera-Agnostic Monocular 3D Reconstruction with Composite Ray Fields

Monocular models recover 3D geometry from a single image in one pass, but most assume a pinhole camera and produce distorted geometry on the wide-angle imagery produced by fisheye and panoramic lenses. Camera-aware models remove this assumption by predicting a per-pixel ray field. However, they represent different camera geometries using a single fixed basis to capture both the overall camera projection and its finer variations. We argue that camera geometry can be viewed as having a hierarchy of complexity for representing rays. We present OmniRay, a feed-forward model whose Ray Module is composed of three components: a monotone radial base that captures radially symmetric structure, a spherical-harmonic residual for smooth deviations from it, and a free-form residual for remaining local variation. These components are composed as successive rotations on the sphere, allowing the representation to accommodate diverse camera geometries. Given a single image, OmniRay predicts a ray direction and an up-to-scale radial distance for every pixel, from which a 3D point cloud is recovered without needing camera information at inference time. Extensive experiments across different camera configurations demonstrate that OmniRay establishes a new state-of-the-art in camera-agnostic monocular 3D reconstruction.
Oct 22, 2026

HiP3R: Hierarchical Pose Graphs for Long-Horizon 3D Reconstruction

Feed-forward models accurately estimate camera motion and 3D geometry from short video windows. However, computational and memory costs prevent direct inference on sequences containing thousands of frames. Long videos must therefore be processed as overlapping chunks. This division produces independent local reconstructions with inconsistent coordinate frames and duplicate estimates of shared cameras. Achieving global consistency requires propagating constraints from adjacent overlaps and distant revisits across the sequence. This requirement naturally motivates pose graph optimization. However, the choice of graph level creates a trade-off. Chunk-level graphs efficiently resolve scale drift and close large loops. Yet, their rigid transformations cannot correct internal errors or conflicting frame predictions. Conversely, frame-level optimization offers local flexibility, but scaling it globally is computationally expensive and can break metric scale. HiP3R resolves this trade-off using a hierarchical pose graph. HiP3R first optimizes a robust chunk-level Sim(3)\mathrm{Sim}(3) graph to establish the global layout and scale. It then holds this layout fixed and estimates one shared SE(3)\mathrm{SE}(3) camera per frame, turning overlapping window predictions into a single trajectory. Experiments across multiple benchmarks demonstrate the effectiveness of HiP3R in improving camera trajectories and 3D reconstruction. Our method applies to both chunk-based and streaming backends without additional training or fine-tuning. The codes are available at https://kkakkkka.github.io/HiP3R.
Oct 22, 2026

DRAFT: Decoding and Reconstructing 3D Room Layouts from fMRI and Camera Trajectories

While functional magnetic resonance imaging (fMRI) decoders reconstruct images and 3D objects, recovering the metric geometry of rooms from a moving viewpoint remains an open challenge. We introduce, to our knowledge for the first time, the task of reconstructing 3D layouts of held-out scenes using only the known camera trajectories and the fMRI signals of participants watching indoor walkthroughs, without any visual frames. The task is hard because of the temporal mismatch between rapid visual changes and blurred fMRI signals. Our method, DRAFT (Decoding and Reconstructing 3D room lAyouts from fMRI and Trajectories), extracts a compact layout representation from the early visual cortex and fuses this neural code with the geometric free space derived from camera motion. To evaluate it, we propose layout r, the Pearson correlation between predicted and ray-cast log depths. On ten held-out scenes (20 participants), DRAFT reaches a layout r of 0.480, outperforming the camera path alone (0.446) and seven fMRI decoders (at most 0.429). DRAFT also works as a plug-in that improves all seven decoders without retraining, raising their layout r from 0.171-0.429 to 0.487-0.493. Brain activity and camera motion can thus recover the metric layout of unseen rooms, a first step toward reconstructing the 3D spaces people see.
Oct 22, 2026

H-2DGS: Height-Field Gaussian Splatting for Sharp and Deformable Neurosurgical Surface Reconstruction

Neurosurgical navigation commonly presents magnetic resonance imaging (MRI) as 2D slices, making it difficult to relate the anatomy in those slices to the cortical surface observed through the surgical microscope. We seek a common 3D visualization that combines a photorealistic reconstruction of the exposed cortex during tumor surgery with whole-brain MRI context and diffusion tractography pathways. This requires the reconstructed cortex to form a clear boundary around the brain anatomy rather than a diffuse volume that obscures internal structures. Conventional 3D Gaussian Splatting (3DGS) can reproduce observed views but, under limited surgical view coverage, may distribute opacity through a thick and geometrically ambiguous band. We introduce height-field 2D Gaussian Splatting (H-2DGS), a three-stage pipeline that concentrates a robust 3DGS reconstruction, fuses its depth into a shared truncated signed distance field (TSDF) surface, and represents that surface with curved, surface-oriented Gaussian primitives. Separate height, color, and opacity bases provide local geometry and appearance, while a neighbor-transition objective coordinates overlapping patches. The explicit surface further supports postoperative alignment: MRI-derived depth deforms the reconstruction, and tracked-color supervision with alternating optimization preserves its appearance. Across five real neurosurgical cases, H-2DGS gives low multi-view depth inconsistency, a compact 7.97MB representation, and interactive rendering at 29--42ms per frame. The resulting sharp, deformable surface integrates clearly with tractography and MRI anatomy, including a postoperative resection cavity.
Oct 21, 2026

Plausible but Unimplantable: A Trustworthiness Audit of Shape Completion Foundation Models for Cranial Defect Reconstruction

A patient who has lost part of the skull vault needs an implant that fits the defect exactly, and such implants are still drawn by hand in CAD by specialists, which makes each case slow and expensive. The task can be considered as a shape completion problem, so pretrained completion models, and now the first shape completion foundation models, look like a way to remove that manual step. We audited three publicly released pretrained shape completion models, two of them foundation models, applied zero-shot to 239 cranial defects, some carved from public skull datasets and some real post-craniectomy cases. Each output was compared with simply returning the patient's own remaining bone, and tested for whether the predicted material could be placed without intersecting the bone that is still there. On the 210 carved defects none of the three improved on doing nothing, and on the 29 real cases only one did so beyond chance. Yet not one output, from any model in any configuration, was implantable: none managed both to stay clear of the bone the patient still has and to place more material than doing nothing, where a task-specific volumetric model, trained for this task and selected on these cases, was implantable on all but one. The output that most convincingly resembles a skull is no more placeable than the rest, and at these point densities implant thickness cannot be read off the emitted points at all, which is a property of the output representation rather than of any one model. On this evidence the shape completion foundation models we could obtain cannot be trusted and, used as released, are not usable in clinical practice. Substantial research towards trustworthy foundation models for shape completion is still needed.
Oct 20, 2026

SceneRig: An Agentic System for Simulation-Ready 3D Scene Reconstruction from Single Images

Simulation is a powerful tool for robotics, but building diverse, simulation-ready scenes at scale remains difficult. Reconstructing scenes from single images is a promising alternative, but visual similarity alone is insufficient: objects must also be recovered, accurately placed, and physically grounded. Existing methods that rely on post-hoc physics settling often misplace objects, producing scenes that look correct but behave poorly under manipulation. We present SceneRig, an agentic system that turns a single image into an editable, simulation-ready scene. SceneRig combines geometric, visual, and physical tools to reconstruct a faithful 3D scene. An agent then performs edit-and-validate loops inside a physics simulator to achieve physical stability. Across diverse inputs, SceneRig improves instance segmentation, object-centric appearance, and 3D placement while keeping the reconstructed scenes physically stable. Furthermore, SceneRig is effective for downstream robotics: its scenes predict real-robot policy outcomes more accurately than the baseline method and more reliably support replay of recorded manipulations without trajectory adaptation.