dual-representation integration

This work addresses the common failure of geometric approximation during mesh generation from parametric boundary representation (B-Rep) models, which often corrupts the original topological structure and leads to incorrect adjacency relationships. The paper proposes a topology-first meshing approach that, for the first time, enforces B-Rep topology as a hard constraint at the algorithmic level, rigorously preserving topological invariance throughout the discretization process. Geometric deviation is controlled solely through user-defined tolerances, thereby decoupling topological correctness from geometric approximation. The method produces robust, topologically consistent meshes without requiring post-processing and has been validated on thousands of real-world CAD models—including cases where conventional tools fail—demonstrating significantly improved reliability for downstream applications.

This work addresses the challenge of efficiently integrating dynamic 3D reconstruction with physics simulation, which is hindered by the complexity of collision detection under changing mesh topologies. The authors propose a dual-representation framework that employs a fixed-topology mesh to enable efficient physical simulation while leveraging Gaussian splatting for high-quality rendering. To handle topological changes, they introduce strategies including vertex buffer updates, temporal correspondence tracking, and stencil projection. Their systematic evaluation—the first of its kind—demonstrates a 4.65× speedup in simulation compared to variable-topology baselines, albeit at the cost of a 65–80% reduction in geometric fidelity during topology transitions. These findings reveal a fundamental trade-off between high-fidelity reconstruction and physics-compatible mesh topologies.

This work proposes a novel paradigm for boundary representation (B-Rep) generation based on composable $k$-cell particles to address the challenges of weak context awareness and poor error recovery arising from the tight coupling between geometry and topology and the complex hierarchical structure in conventional approaches. By leveraging shared latent representations of boundaries, the method achieves geometric coupling while decoupling the traditional hierarchy, enabling unified joint generation of vertices, edges, and faces and eliminating cascading dependencies. Integrated with a multimodal flow-matching framework and explicit local particle representations, the approach supports both unconditional generation and conditional 3D reconstruction from single-view images or point clouds. The resulting CAD models demonstrate superior fidelity, validity, editability, and performance in downstream tasks such as local repair compared to existing methods.

This work addresses the reliance on global synchronization and collective communication in distributed anisotropic mesh adaptation by proposing a novel decoupled approach that separates mesh generation from performance optimization. The method first processes subdomain boundaries on shared-memory multicore nodes, then distributes subdomains across an HPC cluster for parallel interior mesh generation, while freezing adapted boundaries to ensure consistency. A key innovation is the introduction of a semi-speculative execution model, integrated with a cc-NUMA-aware shared-memory mesh generator, a distributed runtime system, and a boundary-freezing strategy, which collectively eliminate the need for global synchronization. Experimental results demonstrate that the approach efficiently generates high-quality meshes with nearly one billion elements, achieving scalability and performance comparable to state-of-the-art HPC meshing software.

Single-image 3D scene generation with multiple objects faces severe challenges including heavy occlusion and object coupling-induced geometric distortion. To address these, we propose a two-stage differentiable framework: first, leveraging off-the-shelf image-to-3D models to independently reconstruct per-object meshes; second, jointly optimizing global scene layout via differentiable rendering, incorporating a novel optimal transport-driven long-range appearance loss and a high-level semantic loss in a synergistic constraint mechanism—enabling unified modeling of object-level geometric independence and scene-level structural consistency. Our approach integrates differentiable rendering, optimal transport theory, and semantics-guided gradient optimization. Evaluated on multi-object benchmarks, our method significantly improves geometric detail fidelity, object separation, and global coherence, outperforming state-of-the-art single-image 3D generation methods both quantitatively and qualitatively.

为解决3D对象生成中缺乏部件级精确控制的问题，MultiCube通过两阶段扩散过程和独立部件条件编码方法实现对各部分语义及空间布局的精细控制。

Traditional rendering approaches face significant challenges when handling complex non-fractal geometries, including high memory overhead, limited topological expressiveness, and inflexible animation control. This work proposes a general-purpose rendering framework based on composable function systems, leveraging GPU-accelerated mesh-free representations to enable efficient generation and manipulation of intricate objects. The framework introduces Quibble, a metaprogramming system that facilitates dynamic composition of functional components. By extending function systems beyond conventional rendering into broader visualization and simulation domains, the method supports modeling of topologically non-trivial structures, controllable in-between frame animation, and point-cloud deformation, while maintaining strong interoperability. Experimental results demonstrate that the framework achieves both high performance and expressive artistic control across diverse tasks, including image synthesis, animation authoring, and geometric simulation.

Existing 3D Gaussian Splatting-based methods for deformable object reconstruction lack explicit surface topology, making it difficult to preserve sharp part boundaries and motion consistency. This work proposes a method to reconstruct articulated, connected triangle meshes with part-wise rigid motions from multi-view images by jointly optimizing a dynamic mesh field within a mesh-native differentiable rendering framework. The approach introduces bidirectional motion consistency constraints at both vertex and pixel levels and features a novel part-aware constrained Delaunay remeshing strategy that aligns mesh topology with semantic parts. Evaluated on the newly introduced Articulate-100 benchmark, the method significantly outperforms existing 3DGS approaches in joint parameter estimation and part-level geometry reconstruction, particularly excelling on objects with multiple moving components.

Existing autoregressive methods for 3D mesh generation suffer from slow inference and error accumulation, making them ill-suited for real-time creation of high-quality meshes with artist-friendly topology. This work proposes the first native mesh generation framework based on flow matching: it employs a vertex-set mesh VAE to encode meshes into continuous latent representations and introduces a two-stage cascaded flow matching model that enables single-pass, parallel decoding from image to complete mesh. The approach eliminates the need for vertex quantization and welding, supports interactive generation (median time of 6 seconds), explicit control over face count, and multi-part asset modeling. It significantly outperforms current autoregressive baselines in both geometric fidelity and generation efficiency, achieving speedups of over an order of magnitude.

Manually editing heterogeneous collision meshes in bulk is time-consuming and poorly scalable, while existing automatic methods often fail to accurately capture user intent. This work introduces neural symbolic program synthesis to 3D collision mesh editing for the first time, formulating the task as a programming-by-example problem: users provide only a few edited examples, and the system automatically synthesizes a reusable program that generalizes to similar meshes. Evaluated on 24 tasks involving 600 meshes, the approach successfully completes 23 tasks, requiring an average of just 2.2 examples per task and synthesizing programs in approximately 3.5 seconds, thereby significantly improving both editing efficiency and scalability.