With recent advances in VR, 3D sketches have emerged as a powerful medium for 3D model creation. However, while they already provide the user with a good perception of the intended shape, they must be surfaced before any reuse in a downstream application. This remains a challenge when sketches are unoriented, i.e., when they are simply sets of unsorted 3D strokes, without any additional normal information. We introduce NeuralSketch2Surf, the first fast and robust neural surfacing solution that processes arbitrarily unoriented sketches at interactive rates. Our approach uses S2V-Net, a new transformer network designed to mesh 3D sketches. Instead of directly inferring complex functions to represent shapes, we focus on predicting an occupancy grid, then refined using a custom smoothing function to create the desired surface. Thanks to a lightweight architecture that enables fast predictions, our method produces results in less than 2 seconds, in contrast to SOTA techniques that can take minutes or even hours. Extensive evaluations demonstrate that our method is not only fast but also generates closed surfaces with high geometric, topological, and perceptual accuracy.
论文检索
输入标题、作者或关键词，从 17,487 篇学术成果中精准定位
Normal orientation has been studied extensively, yet reliably orienting normals on scene-scale point clouds remains challenging. Most existing approaches are tailored to watertight surfaces and often degrade on non-watertight scenes. The few methods that do support non-watertight geometry are typically too slow to handle very large inputs. In this paper, we propose a robust and efficient approach for orienting normals on super-large, non-watertight scenes. Our key insight is that, although global shapes vary widely, local geometric structures are often similar. Based on this observation, we adopt a divide-and-conquer strategy that decomposes the input into smaller patches. We then train a self-conditioning neural network to orient normals within each patch, enabling high efficiency and straightforward parallelization. The remaining challenge is to enforce global consistency across patches. To this end, we train a second network to predict pairwise consistency weights for adjacent patches under potential flipping. Using these weights, we cast global patch reconciliation as a 0-1 integer programming with <1000 binary variables, which can be solved efficiently with off-the-shelf solvers. Since both networks operate only on local patches, our method generalizes well across datasets. Experiments on diverse scene-level data demonstrate that our approach is robust and scales to inputs with up to 50 million points. Code and pre-trained models are publicly available at https://github.com/zd-lee/MegaNorm.
Radar neural reconstruction methods have recently achieved robust 3D scene occupancy from radar measurements alone, as they provide metric depth and are insensitive to adverse weather and low light. However, while these methods can recover some 3D geometry, their input radar data mixes information across elevation into a 2D range-azimuth measurement. This fundamentally limits their elevation resolution, especially in automotive scenes with limited vertical baselines. Camera images offer the opposite tradeoff: they contain strong, high resolution cues for object elevation but struggle with accurate depth and in adverse conditions. We propose CaRaFe, a method that leverages the complementary strengths of camera and radar for 3D reconstruction in challenging urban settings. CaRaFe employs a single multi-modal neural field, relying on conventional novel view synthesis as a supervision signal. Radar supervision provides a valuable geometric constraint to camera rendering that reduces shape-radiance ambiguity, while camera supervision allows for more accurate object elevation disambiguation and fewer missing structures in regions weakly observed by radar. We validate CaRaFe across diverse in-the-wild driving scenes, demonstrating favorable reconstruction quality over both radar and camera methods. Code for this paper is available at light.princeton.edu/carafe.
Surface reconstruction from unoriented point clouds is a fundamental problem in geometry processing. While recent winding-number-based approaches, such as Parametric Gauss Reconstruction (PGR) and Anisotropic Gauss Reconstruction (AGR), provide elegant implicit formulations, they often suffer from underdetermined or ill-conditioned linear systems. In this work, we analyze the theoretical root of this instability: the winding number formula is oblivious to a broad family of tangential perturbations of the normal field, which introduces a non-trivial nullspace in the resulting integral equations. To address this, we present G3R, a simple yet effective method based on Green’s third identity for harmonic functions. This approach eliminates the tangential nullspace with a basis of harmonic functions, resulting in a well-conditioned linear system which can be solved directly without additional regularization. Furthermore, we apply a Gaussian mollification of the fundamental solution, which handles near-field singularities robustly. We adopt a matrix-free implementation and accelerate it with a hierarchical scheme to scale our method to large point clouds. Experiments demonstrate that our approach achieves highly competitive performance in both normal orientation accuracy and surface reconstruction quality. Our code is publicly available at https://github.com/ne0-wu/G3R.
The fundamental limitation of traditional strand-based modeling is not simply data scarcity, but the ill-posedness of inferring complex 3D fields from 2D imagery without structural constraints. This unconstrained regression leads to catastrophic failures in resolving both global occlusion (e.g., in ponytails) and local directionality (e.g., in curls), resulting in over-smoothed, plausible-but-incorrect geometries. To resolve this, we integrate the strong geometric priors of Large Reconstruction Models (LRMs) into the strand generation pipeline. Using the LRM mesh as a structural anchor, we employ a novel Dual Orientation AutoEncoder to lift coarse geometry into high-fidelity strands. By resolving vector field singularities through latent-space optimization and surface-guided refinement, our method effectively disentangles complex topological structures, setting a new benchmark for robustness and accuracy in hair reconstruction.
Motion warping is a core technique in character animation that enables the adaptation of existing motion data to novel spatio-temporal constraints. Conventional motion warping methods often rely on heuristic modifications that can violate physical consistency or introduce visual artifacts. More recent learning-based editing approaches improve realism, but many of them encode motion into tightly entangled latent space, which makes them struggle to balance editing flexibility and content preservation. To address this, we propose a novel deep motion warping framework that explicitly disentangles the motion structure from global and stylistic attributes for intuitive motion editing. Our key insight is to leverage learned phase features as a continuous and robust representation of the underlying structure, and explicitly disentangle motion into root velocity, phase, and learned latent variables using a phase-conditioned diffusion autoencoder. This design supports a wide range of editing operations, including root motion warping, motion exaggeration, time warping, and style transfer by directly manipulating decoupled components, without requiring paired training data. Extensive experiments demonstrate that our approach enables high-level, flexible motion editing while strictly preserving the structural consistency and physical plausibility of the source motion
Developing controllers capable of completing a wide range of tasks in a natural and life-like manner is a key challenge in enabling practical applications of physics-based character animation. In this work, we introduce Generative Pretrained Controllers (GPC), which leverage tokenization and next-token modeling to create general-purpose, reusable generative controllers from large-scale motion datasets. Our framework utilizes end-to-end reinforcement learning to jointly optimize a "motion vocabulary", modeled via Finite Scalar Quantization (FSQ), along with a corresponding control policy that can map the discrete codes to physics-based controls. After the "codebook" has been learned, the underlying structure of this large vocabulary is modeled by training a GPT-style autoregressive transformer, leading to a powerful generative controller that generates controls for a physically simulated character by performing next-token prediction. Once the generative controller has been trained, we propose a suite of adaptation techniques for finetuning the controller for new downstream tasks. Our proposed framework greatly simplifies the training process compared to previous tokenized methods, and achieves a 99.98% success rate in reproducing a vast corpus of motion clips. The generative controller exhibits a variety of natural emergent behaviors, such as responsive behaviors to perturbations and recovery behaviors after falling. This results in highly robust general purpose controllers for a variety of downstream applications.
We propose a robust texture optimization framework that handles inconsistent AI-generated targets by operating directly within non-differentiable production pipelines. Standard optimization approaches struggle with inconsistent targets, often producing blurry textures and ghosting artifacts. Moreover, existing inverse rendering methods rely on differentiable renderers, causing a rendering gap when assets are deployed in production engines. To address the inconsistencies, we introduce a robust formulation that jointly optimizes a deformation grid and an uncertainty map. This formulation effectively decouples geometric misalignment from semantic hallucinations. To avoid the rendering gap, we leverage finite-difference gradient estimation to operate entirely within standard rasterizers. Extensive experiments demonstrate that our approach recovers sharp, high-fidelity textures from inconsistent AI-generated targets, achieving higher visual quality than existing methods. This makes our technique a practical tool for film production, gaming, and virtual reality, where flexibility and visual quality are paramount.
Volumetric video (VV) streaming delivers truly immersive viewing experiences over the Internet, serving as a critical foundation for next-generation applications, including immersive telepresence in the metaverse, the surveillance of remote ecological systems, and robotic teleoperation for embodied AI, and beyond. Beyond immersive viewing, these applications turn VV streaming into a real-time interface to remote physical environments, imposing new system-level demands for photorealistic scene representation, low-latency interaction, and robust performance under heterogeneous network conditions. 3D Gaussian Splatting (3DGS) has been widely used for real-time photorealistic rendering, offering superior visual quality and rendering performance, but it faces challenges due to bandwidth consumption. Furthermore, as the foundation of adaptive VV streaming, existing Levels of Detail (LoD) methods based on density are not well-suited to Gaussian representations, leading to visible gaps and severe quality degradation. Recent studies have also explored attribute compression techniques to reduce bandwidth consumption. Our preliminary studies reveal that aggressive attribute compression primarily causes color distortion, which can be effectively corrected in the rendered image using a reference image. Motivated by these findings, we propose a novel Color-Adaptive scheme for adaptive VV streaming that uses vector quantization (VQ) to establish LoDs and correct color distortions with low-resolution reference images. We further present CAGS, an adaptive VV streaming system compatible with diverse Gaussian representations, which integrates the Color-Adaptive scheme by rendering reference images on the streaming server and performing color restoration on the client. Extensive experiments on our prototype system demonstrate that CAGS outperforms the existing adaptive streaming systems in PSNR by 5 ∼ 20 dB under fluctuating bandwidth, operates significantly faster than existing scalable Gaussian compression methods, and generalizes across different Gaussian representations. The code is available at https://github.com/yindaheng98/ColorAdaptiveGaussianSplatting.
Voronoi diagrams, and their more general weighted counterpart, power diagrams, are fundamental geometric constructs with wide-ranging applications in biology, physics simulation, and computer graphics. Recently, they have gained renewed attention in mesh-based neural rendering. Despite being extensively studied, the construction of 3D Voronoi diagrams for large-scale point sets remains computationally expensive, limiting their adoption in large-scale applications. Existing CPU-based approaches typically rely on computing its dual, the Delaunay tetrahedralization, but are prohibitively slow for large diagrams, while GPU-based methods either struggle to scale efficiently to large point sets or assume homogeneous point distributions. The weighted case, power diagrams, is even less explored in this context. Existing approaches are typically tailored to the application at hand, assuming homogeneous point distributions and small weight variations, making them unsuitable for general use in more complex heterogeneous data. In this paper, we present a highly parallelizable GPU algorithm for the fast construction of large-scale 3D Voronoi and power diagrams. Our approach constructs each convex cell from a weighted 3D point by progressively clipping an initial cell volume against bisecting planes induced by candidate neighboring points. To efficiently identify candidate neighbors under arbitrary spatial distributions, we introduce a culling criterion based on directional geometric bounds of the evolving cell, combined with a hierarchical best-first traversal of bounding volumes. We achieve performance on par with state-of-the-art Delaunay tetrahedralization methods on small and moderate problem sizes, while exhibiting robust scalability to large point sets and diverse spatial distributions. Moreover, our method naturally generalizes to power diagrams without additional assumptions. To facilitate reproducibility and future research, we release our source code, see https://research.zenseact.com/publications/paragram.
2D constrained Delaunay triangulation (CDT) is a key component in CAD, visualization, and scientific computing. We present a highly parallel GPU-based method capable of computing CDTs on inputs with millions of vertices, while efficiently and robustly enforcing arbitrary valid constraints. On benchmarks with complex constraints, our method typically delivers several-fold speedups over the prior state-of-the-art GPU approach and is roughly an order of magnitude faster than widely used CPU implementations. The efficiency stems from an improved parallel edge-flipping scheme and a constraint-handling algorithm with predictable parallel behavior, enabling stable performance even on adversarial and difficult configurations.
3D Gaussian Splatting has emerged as a powerful scene representation for real-time novel-view synthesis. However, its standard adaptive density control relies on screen-space positional gradients, which do not distinguish between geometric misplacement and frequency aliasing, often leading to either over-blurred high-frequency textures or inefficient over-densification. We present a structure-aware densification framework. Our key insight is that the decision to subdivide a Gaussian should be driven by an explicit comparison between its projected screen-space extent and the local structure of the texture it seeks to represent. We introduce a multi-scale frequency analysis combining structure tensors with Laplacian scale space analysis to estimate the dominant frequency at each pixel, enabling robust supervision across varying texture scales. Based on this analysis, we define η , a per-Gaussian, per-axis frequency violation metric that indicates when a primitive may be under-resolving local texture details. Unlike methods that perform isotropic splitting (e.g., splitting each Gaussian into two smaller ones with uniform shape), our approach performs anisotropic splitting. For each axis with high η , we compute a split factor to better resolve the local frequency content. We further introduce a multiview consistency criterion that aggregates η observations across multiple views. By performing densification early and faster, we skip the lengthy iterative densification phases required by baseline methods and achieve significantly faster convergence. Experiments on standard benchmarks demonstrate that our method also achieves superior reconstruction quality, particularly in high-frequency regions.
Advances in large language models (LLMs) have sparked interest in automating parametric CAD modeling through natural language. Existing LLM-based approaches often treat CAD modeling as flat text generation, overlooking the hierarchical structure and geometric constraints inherent in CAD programs. We present CAD-Factory, a Text-to-CAD generation system for language-driven CAD modeling that explicitly models the structural and parametric semantics of CAD programs. Our core contribution is a new formulation of CAD generation as structured program synthesis, coupled with a learnable hierarchical CAD program representation that disentangles structural topology from parameter instantiation. Building on this formulation, CAD-Factory adopts a manager–programmer–reviewer architecture: a planning agent infers program hierarchy, a coding agent instantiates symbolic and numerical parameters, and an evaluation module enforces structural validity and geometric feasibility, which supports structure-aware reasoning, constraint-consistent generation, and interpretable program synthesis. An editor–viewer loop further enables part-aware code refinement through visual feedback, supporting iterative and controllable design workflows. We also contribute a process-oriented annotation pipeline and a Text–CAD dataset with parameter-free or partially specified descriptions that reflect real-world design expression. Extensive experiments demonstrate that CAD-Factory significantly improves structural correctness and geometric consistency across diverse CAD code generation and editing tasks, establishing a structured, interpretable, and robust framework for AI-assisted design.
Boundary Representation (BRep) is the standard format for Computer-Aided Design (CAD), yet reconstructing high-quality BReps from single-view images remains challenging due to the complexity of topological constraints and operation sequences. We present Img2CADSeq, a multi-stage pipeline that overcomes these limitations by encoding CAD sequences into a three-level hierarchical codebook. Guided by an importance prioritization, this strategy values profiles over details, compressing long sequences into a stable discrete latent space. To bridge the modality gap, we leverage a coarse-to-fine point cloud intermediate, aligning 2D visual features with 3D CAD sequences via contrastive learning to condition a VQ-Diffusion model. Supported by newly introduced CAD-220K and PrintCAD datasets, our approach ensures robust industrial domain adaptation. Extensive experiments demonstrate that Img2CADSeq significantly outperforms state-of-the-art methods, producing standard STEP files that can be directly used in commercial CAD software. Code and data for this paper are at https://github.com/Rilpraa0110/Img2CADSeq
Generative CAD modeling has broad design and application potential. Despite significant advances in Boundary Representation (B-Rep) generation, the dominant representation in CAD, existing methods largely depend on uniformly sampled point- or grid-based geometry representations, sacrificing native surface types and parameters and thereby limiting geometric fidelity and downstream usability. We present ParaCAD, an autoregressive framework for point-cloud-conditioned B-Rep generation that directly operates on native parametric surfaces. ParaCAD introduces a surface-centric tokenization that explicitly encodes each face by its exact surface type and continuous parameters, preserving the intrinsic semantics of CAD geometry. Our model first generates parametric surfaces with constrained UV domains, and then constructs a valid B-Rep by globally intersecting these surfaces to recover edges and vertices. ParaCAD places point-cloud-conditioned generation at the core of B-Rep synthesis, making it practical for user-guided reconstruction and seamless integration into existing 3D generation pipelines. Extensive experiments demonstrate that ParaCAD produces accurate B-Reps with faithful point-cloud alignment, outperforming point-based baselines in geometric precision, robustness, watertightness and downstream usability.
Boundary Representation (B-rep) is the most commonly used data format in Computer-Aided Design (CAD) due to its analytical precision and direct support for parametric editing. However, its heterogeneous data structure— continuous parametric geometry with discrete topological graphs—poses fundamental challenges for deep learning models. Existing methods often directly predict the heterogeneous B-rep graph, relying on fixed-size padding or sequential tokenization to handle the varying cardinality of the geometric primitives. These approaches struggle with the combinatorial complexity of CAD models. The discrete, non-differentiable nature of the graph data structure prevents end-to-end optimization of the geometry and watertightness. In this work we introduce DualBrep, a novel continuous representation that unifies B-rep geometry and topology within a fully structured Euclidean domain. DualBrep encodes a CAD model using dual scalar fields: a Signed Distance Function (SDF) to represent the global shape geometry, and an Unsigned Distance Field (UDF) that implicitly encodes the topological structure via a Voronoi partitioning of the surface elements. Rather than processing these fields independently, we compress them into a single latent space. While the dual-field formulation alone already gives reconstruction a flexible, primitive-free segmentation signal that adapts to arbitrary face counts and surface types, the shared latent also makes generation tractable: a Flow Matching model can sample geometry and topology jointly from a single code, avoiding the error accumulation that plagues sequential or autoregressive B-rep predictors. Finally, we use a neural rebuilder to extract explicit B-rep models—comprising both prismatic and free-form primitives—directly from our continuous dual scalar fields. We demonstrate that DualBrep serves as a robust, unified backbone for CAD B-rep learning, achieving strong performance in both reverse engineering from raw point clouds and generative modeling via latent flow matching. Code is available at https://github.com/AutodeskAILab/DualBrep.
Inverse sculpting modeling aims to decompose a sculpted mesh into an underlying base shape and reusable geometric details, enabling non-expert users to inherit professional sculpting effort. We present InvSculpt, a novel inverse sculpting framework that decomposes a sculpted mesh into a high-fidelity underlying shape and reusable geometric details represented as a vector displacement field (VDF). Our approach combines semantic priors from text-guided 2D image editing with a 3D rectified flow model to perform inversion-based, mask-free detail removal, recovering an underlying shape that preserves the identity of the source mesh. To represent sculpted details in a lossless and transferable manner, we extract a VDF defined on the surface of the recovered underlying shape and learn a continuous neural representation for geometry-aware transfer. We observe that standard conditional sampling after inversion often suffers from trajectory drift, leading to identity shift and low-frequency distortion. To address this issue, we introduce a trajectory correction strategy that constrains early sampling steps to follow the inversion path, effectively stabilizing subsequent conditional guidance. This design enables robust detail removal and precise extraction of the VDF. Extensive experiments demonstrate that InvSculpt achieves significantly higher-quality mesh decomposition than prior methods and supports a wide range of applications, including geometry redesign and high-fidelity geometric detail transfer.
The eigendecomposition of the Laplace–Beltrami Operator (LBO) is fundamental to geometric analysis, yet computing its low-frequency eigenmodes remains a significant bottleneck due to the high cost of iterative solvers on large-scale data. To amortize this cost, we introduce the Neural Eigenspace Operator (NEO), a feed-forward framework designed to predict the spectrum directly from point clouds. Crucially, NEO circumvents the ill-posed nature of standard eigenvector regression, which suffers from intrinsic sign flips and rotation ambiguities, by learning the stable, invariant low-frequency subspace instead. Specifically, the network predicts a redundant set of basis functions whose span robustly covers the target eigenspace, allowing for the recovery of accurate eigenpairs via a lightweight Rayleigh–Ritz refinement. To handle irregular sampling, we propose a mass-aware neural operator that incorporates per-point area weights into attention-based aggregation, improving robustness to non-uniform densities and enabling zero-shot generalization across resolutions. Our approach achieves near-linear runtime scaling and substantial wall-clock speedups over iterative solvers at comparable accuracy, and exhibits strong zero-shot transfer to high-resolution point clouds. The resulting eigenpairs support standard spectral geometry tasks, while the raw basis functions provide effective point-wise features for downstream learning. Code: https://github.com/Adversarr/NEO.
Cultural context profoundly shapes how people interpret online content, yet vision–language models (VLMs) remain predominantly trained through Western or English-centric lenses. This limits their fairness and cross-cultural robustness in tasks like hateful meme detection. We introduce a systematic evaluation framework designed to diagnose and quantify the cross-cultural robustness of state-of-the-art VLMs across multilingual meme datasets, analyzing three axes: (i) learning strategy (zero-shot vs. one-shot), (ii) prompting language (native vs. English), and (iii) translation effects on meaning and detection. Results show that the common ''translate-then-detect'' approach deteriorate performance, while culturally aligned interventions — native-language prompting and one-shot learning — significantly enhance detection. Our findings reveal systematic convergence toward Western safety norms and provide actionable strategies to mitigate such bias, guiding the design of globally robust multimodal moderation systems.
A lesson plan (LP) is a structured guide outlining instructional objectives, methods, and assessments to ensure organized learning. However, existing LP creations are often time-consuming, inconsistent in structure, and lack pedagogical mechanisms for real-time adaptation to diverse learner needs. To address these issues, we propose a co llaborative multi-role a gent framework called COMA for automatic LP generation. COMA formulates LP generation as a collaborative workflow among multiple LLM agents with distinct pedagogical expertise: (1) the novice agent that represents a novice teacher possesses an overarching understanding of the intended lesson flow but demonstrates limited precision in implementing the specific instructional actions; (2) the veteran agent that represents an experienced teacher demonstrates deep familiarity with the curriculum, textbooks, and the knowledge components embedded in each unit; and (3) the master agent that represents a pedagogical expert exhibits a well-developed and confident grasp of lesson progression, with the ability to design, adapt, and implement specific instructional actions effectively and responsively. Through an iterative workflow, these agents collaboratively refine LP quality. Comprehensive experiments across five subjects, using expert-designed metrics, demonstrate that COMA significantly outperforms state-of-the-art methods, producing lesson plans with superior quality, coherence, and pedagogical alignment. Our framework offers a robust solution for generating deployable instructional content at scale. Data and code are available at https://github.com/ai4ed/COMA-LessonPlan.