Zixiong Wang

2026

1. CGF

   

   HiMat: DiT-based Ultra-High Resolution SVBRDF GenerationFirst Author

   Zixiong Wang, Jian Yang, Yiwei Hu, Miloš Hašan, and Beibei Wang

   Computer Graphics Forum (Proc. Eurographics), Mar 2026

   Abs Paper arXiv Bib Project Page

   Creating ultra-high-resolution spatially varying bidirectional reflectance functions (SVBRDFs) is critical for photorealistic 3D content creation, to faithfully represent fine-scale surface details required for close-up rendering. However, achieving 4K generation faces two key challenges: (1) the need to synthesize multiple reflectance maps at full resolution, which multiplies the pixel budget and imposes prohibitive memory and computational cost, and (2) the requirement to maintain strong pixel-level alignment across maps at 4K, which is particularly difficult when adapting pretrained models designed for the RGB image domain. We introduce HiMat, a diffusion-based framework tailored for efficient and diverse 4K SVBRDF generation. To address the first challenge, HiMat performs generation in a high-compression latent space via DC-AE, and employs a pretrained diffusion transformer with linear attention to improve per-map efficiency. To address the second challenge, we propose CrossStitch, a lightweight convolutional module that enforces cross-map consistency without incurring the cost of global attention. Our experiments show that HiMat achieves high-fidelity 4K SVBRDF generation with superior efficiency, structural consistency, and diversity compared to prior methods. Beyond materials, our framework also generalizes to related applications such as intrinsic decomposition.

   @article{wang2026himat,
     title = {HiMat: DiT-based Ultra-High Resolution SVBRDF Generation},
     author = {Wang, Zixiong and Yang, Jian and Hu, Yiwei and Ha{\v{s}}an, Milo{\v{s}} and Wang, Beibei},
     journal = {Computer Graphics Forum (Proc. Eurographics)},
     year = {2026},
     month = mar,
     publisher = {Wiley},
     doi = {10.1111/cgf.70343},
   }

2. CVPR

   

   FabricGen: Microstructure-Aware Woven Fabric Generation

   Yingjie Tang, Di Luo, Zixiong Wang, Xiaoli Ling, Jian Yang, and Beibei Wang

   In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2026

   Abs arXiv Bib

   Woven fabric materials are widely used in rendering applications, yet designing realistic examples typically involves multiple stages, requiring expertise in weaving principles and texture authoring. Recent advances have explored diffusion models to streamline this process; however, pre-trained diffusion models often struggle to generate intricate yarn-level details that conform to weaving rules. To address this, we present FabricGen, an end-to-end framework for generating high-quality woven fabric materials from textual descriptions. A key insight of our method is the decomposition of macro-scale textures and micro-scale weaving patterns. To generate macro-scale textures free from microstructures, we fine-tune pre-trained diffusion models on a collected dataset of microstructure-free fabrics. As for micro-scale weaving patterns, we develop an enhanced procedural geometric model capable of synthesizing natural yarn-level geometry with yarn sliding and flyaway fibers. The procedural model is driven by a specialized large language model, WeavingLLM, which is fine-tuned on an annotated dataset of formatted weaving drafts, and prompt-tuned with domain-specific fabric expertise.

   @inproceedings{tang2026fabricgen,
     title = {FabricGen: Microstructure-Aware Woven Fabric Generation},
     author = {Tang, Yingjie and Luo, Di and Wang, Zixiong and Ling, Xiaoli and Yang, Jian and Wang, Beibei},
     booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
     year = {2026},
   }

3. SIGGRAPH

   

   PureSample: Neural Materials Learned by Sampling Microgeometry

   Zixuan Li, Zixiong Wang, Jian Yang, Miloš Hašan, and Beibei Wang

   In ACM SIGGRAPH 2026 Conference Papers, 2026

   Abs arXiv Bib

   Traditional physically-based material models rely on analytically derived bidirectional reflectance distribution functions (BRDFs), typically by considering statistics of micro-primitives such as facets, flakes, or spheres, sometimes combined with multi-bounce interactions such as layering and multiple scattering. We present PureSample: a novel neural BRDF representation that allows learning a material’s behavior purely by sampling forward random walks on the microgeometry. The approach uses two learnable components: first, the sampling distribution is modeled using a flow matching neural network, which allows both importance sampling and pdf evaluation; second, we introduce a view-dependent albedo term, captured by a lightweight neural network, which allows for converting a scalar pdf value to a colored BRDF value for any pair of view and light directions.

   @inproceedings{li2026puresample,
     title = {PureSample: Neural Materials Learned by Sampling Microgeometry},
     author = {Li, Zixuan and Wang, Zixiong and Yang, Jian and Ha{\v{s}}an, Milo{\v{s}} and Wang, Beibei},
     booktitle = {ACM SIGGRAPH 2026 Conference Papers},
     year = {2026},
   }

2024

1. CAGD

   

   A Task-driven Network for Mesh Classification and Semantic Part Segmentation

   Qiujie Dong, Xiaoran Gong, Rui Xu, Zixiong Wang, Junjie Gao, Shuangmin Chen, Shiqing Xin, Changhe Tu, and Wenping Wang

   Computer Aided Geometric Design, 2024

   Abs Paper Bib Code Project Page

   Given the rapid advancements in geometric deep-learning techniques, there has been a dedicated effort to create mesh-based convolutional operators that act as a link between irregular mesh structures and widely adopted backbone networks. Despite the numerous advantages of Convolutional Neural Networks (CNNs) over Multi-Layer Perceptrons (MLPs), mesh-oriented CNNs often require intricate network architectures to tackle irregularities of a triangular mesh. These architectures not only demand that the mesh be manifold and watertight but also impose constraints on the abundance of training samples. In this paper, we note that for specific tasks such as mesh classification and semantic part segmentation, large-scale shape features play a pivotal role. This is in contrast to the realm of shape correspondence, where a comprehensive understanding of 3D shapes necessitates considering both local and global characteristics. Inspired by this key observation, we introduce a task-driven neural network architecture that seamlessly operates in an end-to-end fashion. Our method takes as input mesh vertices equipped with the heat kernel signature (HKS) and dihedral angles between adjacent faces. Notably, we replace the conventional convolutional module, commonly found in ResNet architectures, with MLPs and incorporate Layer Normalization (LN) to facilitate layer-wise normalization. Our approach, with a seemingly straightforward network architecture, demonstrates an accuracy advantage. It exhibits a marginal 0.1% improvement in the mesh classification task and a substantial 1.8% enhancement in the mesh part segmentation task compared to state-of-the-art methodologies. Moreover, as the number of training samples decreases to 1/50 or even 1/100, the accuracy advantage of our approach becomes more pronounced. In summary, our convolution-free network is tailored for specific tasks relying on large-scale shape features and excels in the situation with a limited number of training samples, setting itself apart from state-of-the-art methodologies.

   @article{dong2024taskdriven,
     title = {A Task-driven Network for Mesh Classification and Semantic Part Segmentation},
     author = {Dong, Qiujie and Gong, Xiaoran and Xu, Rui and Wang, Zixiong and Gao, Junjie and Chen, Shuangmin and Xin, Shiqing and Tu, Changhe and Wang, Wenping},
     journal = {Computer Aided Geometric Design},
     year = {2024},
     publisher = {Elsevier},
     doi = {10.1016/j.cagd.2024.102304},
   }

2023

1. NeurIPS

   

   Aligning Gradient and Hessian for Neural Signed Distance FunctionCo-first Author

   Ruian Wang^*, Zixiong Wang^*, Yunxiao Zhang, Shuangmin Chen, Shiqing Xin, Changhe Tu, and Wenping Wang

   In Advances in Neural Information Processing Systems (NeurIPS), 2023

   Abs Bib Paper Project Page

   The Signed Distance Function (SDF), as an implicit surface representation, provides a crucial method for reconstructing a watertight surface from unorganized point clouds. The SDF has a fundamental relationship with the principles of surface vector calculus. Given a smooth surface, there exists a thin-shell space in which the SDF is differentiable everywhere such that the gradient of the SDF is an eigenvector of its Hessian matrix, with a corresponding eigenvalue of zero. In this paper, we introduce a method to directly learn the SDF from point clouds in the absence of normals. Our motivation is grounded in a fundamental observation: aligning the gradient and the Hessian of the SDF provides a more efficient mechanism to govern gradient directions. This, in turn, ensures that gradient changes more accurately reflect the true underlying variations in shape. Extensive experimental results demonstrate its ability to accurately recover the underlying shape while effectively suppressing the presence of ghost geometry.

   @inproceedings{wang2023aligning,
     title = {Aligning Gradient and Hessian for Neural Signed Distance Function},
     author = {Wang, Ruian and Wang, Zixiong and Zhang, Yunxiao and Chen, Shuangmin and Xin, Shiqing and Tu, Changhe and Wang, Wenping},
     booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},
     year = {2023},
   }

2. TOG

   

   Neural-Singular-Hessian: Implicit Neural Representation of Unoriented Point Clouds by Enforcing Singular HessianFirst Author

   Zixiong Wang, Yunxiao Zhang, Rui Xu, Fan Zhang, Pengshuai Wang, Shuangmin Chen, Shiqing Xin, Wenping Wang, and Changhe Tu

   ACM Transactions on Graphics (Proc. SIGGRAPH Asia), 2023

   Abs Paper arXiv Bib Video Code Project Page

   Neural implicit representation is a promising approach for reconstructing surfaces from point clouds. Existing methods combine various regularization terms, such as the Eikonal and Laplacian energy terms, to enforce the learned neural function to possess the properties of a Signed Distance Function (SDF). However, inferring the actual topology and geometry of the underlying surface from poor-quality unoriented point clouds remains challenging. In accordance with Differential Geometry, the Hessian of the SDF is singular for points within the differential thin-shell space surrounding the surface. Our approach enforces the Hessian of the neural implicit function to have a zero determinant for points near the surface. This technique aligns the gradients for a near-surface point and its on-surface projection point, producing a rough but faithful shape within just a few iterations. By annealing the weight of the singular-Hessian term, our approach ultimately produces a high-fidelity reconstruction result.

   @article{wang2023neuralsingularhessian,
     title = {Neural-Singular-Hessian: Implicit Neural Representation of Unoriented Point Clouds by Enforcing Singular Hessian},
     author = {Wang, Zixiong and Zhang, Yunxiao and Xu, Rui and Zhang, Fan and Wang, Pengshuai and Chen, Shuangmin and Xin, Shiqing and Wang, Wenping and Tu, Changhe},
     journal = {ACM Transactions on Graphics (Proc. SIGGRAPH Asia)},
     volume = {42},
     number = {6},
     year = {2023},
     publisher = {ACM},
     doi = {10.1145/3618311},
   }

3. SIGGRAPH Asia

   

   A Hessian-Based Field Deformer for Real-Time Topology-Aware Shape Editing

   Yunxiao Zhang, Zixiong Wang, Zihan Zhao, Rui Xu, Shuangmin Chen, Shiqing Xin, Wenping Wang, and Changhe Tu

   In SIGGRAPH Asia 2023 Conference Papers, 2023

   Abs Paper arXiv Bib

   Shape manipulation is a central research topic in computer graphics. Topology editing, such as breaking apart connections, joining disconnected ends, and filling/opening a topological hole, is generally more challenging than geometry editing. In this paper, we observe that the saddle points of the signed distance function (SDF) provide useful hints for altering surface topology deliberately. Based on this key observation, we parameterize the SDF into a cubic trivariate tensor-product B-spline function F whose saddle points {\boldsymbols_i} can be quickly exhausted based on a subdivision-based root-finding technique coupled with Newton’s method. Users can select one of the candidate points, say \boldsymbols_i, to edit the topology in real time. In implementation, we add a compactly supported B-spline function rooted at \boldsymbols_i, which we call a \textitdeformer in this paper, to F, with its local coordinate system aligning with the three eigenvectors of the Hessian. Combined with ray marching technique, our interactive system operates at 30 FPS. Additionally, our system empowers users to create desired bulges or concavities on the surface. An extensive user study indicates that our system is user-friendly and intuitive to operate. We demonstrate the effectiveness and usefulness of our system in a range of applications, including fixing surface reconstruction errors, artistic work design, 3D medical imaging and simulation, and antiquity restoration. Please refer to the attached video for a demonstration.

   @inproceedings{zhang2023hessianfield,
     title = {A Hessian-Based Field Deformer for Real-Time Topology-Aware Shape Editing},
     author = {Zhang, Yunxiao and Wang, Zixiong and Zhao, Zihan and Xu, Rui and Chen, Shuangmin and Xin, Shiqing and Wang, Wenping and Tu, Changhe},
     booktitle = {SIGGRAPH Asia 2023 Conference Papers},
     year = {2023},
     publisher = {ACM},
     doi = {10.1145/3610548.3618191},
   }

4. TVCG

   

   Neural-IMLS: Self-supervised Implicit Moving Least-Squares Network for Surface ReconstructionFirst Author

   Zixiong Wang, Pengfei Wang, Pengshuai Wang, Qiujie Dong, Junjie Gao, Shuangmin Chen, Shiqing Xin, Changhe Tu, and Wenping Wang

   IEEE Transactions on Visualization and Computer Graphics, 2023

   Abs Paper arXiv Bib Code Project Page

   Surface reconstruction is a challenging task when input point clouds, especially real scans, are noisy and lack normals. Observing that the Multilayer Perceptron (MLP) and the implicit moving least-square function (IMLS) provide a dual representation of the underlying surface, we introduce Neural-IMLS, a novel approach that directly learns a noise-resistant signed distance function (SDF) from unoriented raw point clouds in a self-supervised manner. In particular, IMLS regularizes MLP by providing estimated SDFs near the surface and helps enhance its ability to represent geometric details and sharp features, while MLP regularizes IMLS by providing estimated normals. We prove that at convergence, our neural network produces a faithful SDF whose zero-level set approximates the underlying surface due to the mutual learning mechanism between the MLP and the IMLS.

   @article{wang2023neuralimls,
     title = {Neural-IMLS: Self-supervised Implicit Moving Least-Squares Network for Surface Reconstruction},
     author = {Wang, Zixiong and Wang, Pengfei and Wang, Pengshuai and Dong, Qiujie and Gao, Junjie and Chen, Shuangmin and Xin, Shiqing and Tu, Changhe and Wang, Wenping},
     journal = {IEEE Transactions on Visualization and Computer Graphics},
     year = {2023},
     pages = {1--16},
     doi = {10.1109/TVCG.2023.3284233},
   }

5. TVCG

   

   Laplacian2Mesh: Laplacian-Based Mesh Understanding

   Qiujie Dong, Zixiong Wang, Manyi Li, Junjie Gao, Shuangmin Chen, Zhenyu Shu, Shiqing Xin, Changhe Tu, and Wenping Wang

   IEEE Transactions on Visualization and Computer Graphics, 2023

   Abs arXiv Bib Code Project Page

   Geometric deep learning has sparked a rising interest in computer graphics to perform shape understanding tasks, such as shape classification and semantic segmentation. When the input is a polygonal surface, one has to suffer from the irregular mesh structure. Motivated by the geometric spectral theory, we introduce Laplacian2Mesh, a novel and flexible convolutional neural network (CNN) framework for coping with irregular triangle meshes (vertices may have any valence). By mapping the input mesh surface to the multi-dimensional Laplacian-Beltrami space, Laplacian2Mesh enables one to perform shape analysis tasks directly using the mature CNNs, without the need to deal with the irregular connectivity of the mesh structure. We further define a mesh pooling operation such that the receptive field of the network can be expanded while retaining the original vertex set as well as the connections between them. Besides, we introduce a channel-wise self-attention block to learn the individual importance of feature ingredients. Laplacian2Mesh not only decouples the geometry from the irregular connectivity of the mesh structure but also better captures the global features that are central to shape classification and segmentation. Extensive tests on various datasets demonstrate the effectiveness and efficiency of Laplacian2Mesh, particularly in terms of the capability of being vulnerable to noise to fulfill various learning tasks.

   @article{dong2023laplacian2mesh,
     title = {Laplacian2Mesh: Laplacian-Based Mesh Understanding},
     author = {Dong, Qiujie and Wang, Zixiong and Li, Manyi and Gao, Junjie and Chen, Shuangmin and Shu, Zhenyu and Xin, Shiqing and Tu, Changhe and Wang, Wenping},
     journal = {IEEE Transactions on Visualization and Computer Graphics},
     year = {2023},
   }

2022

1. TOG

   

   RFEPS: Reconstructing Feature-line Equipped Polygonal Surface

   Rui Xu, Zixiong Wang, Zhiyang Dou, Chen Zong, Shiqing Xin, Mingyan Jiang, Tao Ju, and Changhe Tu

   ACM Transactions on Graphics (Proc. SIGGRAPH Asia), 2022

   Abs arXiv Bib Video Code Project Page

   Feature lines are important geometric cues in characterizing the structure of a CAD model. Despite great progress in both explicit reconstruction and implicit reconstruction, it remains a challenging task to reconstruct a polygonal surface equipped with feature lines, especially when the input point cloud is noisy and lacks faithful normal vectors. In this paper, we develop a multistage algorithm, named RFEPS, to address this challenge. The key steps include (1)denoising the point cloud based on the assumption of local planarity, (2)identifying the feature-line zone by optimization of discrete optimal transport, (3)augmenting the point set so that sufficiently many additional points are generated on potential geometry edges, and (4) generating a polygonal surface that interpolates the augmented point set based on restricted power diagram. We demonstrate through extensive experiments that RFEPS, benefiting from the edge-point augmentation and the feature-preserving explicit reconstruction, outperforms state-of-the-art methods in terms of the reconstruction quality, especially in terms of the ability to reconstruct missing feature lines.

   @article{xu2022rfeps,
     title = {RFEPS: Reconstructing Feature-line Equipped Polygonal Surface},
     author = {Xu, Rui and Wang, Zixiong and Dou, Zhiyang and Zong, Chen and Xin, Shiqing and Jiang, Mingyan and Ju, Tao and Tu, Changhe},
     journal = {ACM Transactions on Graphics (Proc. SIGGRAPH Asia)},
     year = {2022},
     publisher = {ACM},
   }

2. TOG

   

   Restricted Delaunay Triangulation for Explicit Surface Reconstruction

   Pengfei Wang, Zixiong Wang, Shiqing Xin, Xifeng Gao, Wenping Wang, and Changhe Tu

   ACM Transactions on Graphics, 2022

   Abs Paper Bib

   The task of explicit surface reconstruction is to generate a surface mesh by interpolating a given point cloud. Explicit surface reconstruction is necessary when the point cloud is required to appear exactly on the surface. However, for a non-perfect input, such as lack of normals, low density, irregular distribution, thin and tiny parts, and high genus, a robust explicit reconstruction method that can generate a high-quality manifold triangulation is missing. The proposed approach starts from an initial simple surface mesh, alternately performing a Filmsticking step and a Sculpting step of the initial mesh, and converges when the surface mesh interpolates all input points (except outliers) and remains stable. The Filmsticking minimizes the geometric distance between the surface mesh and the point cloud through iteratively performing a restricted Voronoi diagram technique on the surface mesh, whereas the Sculpting bootstraps the Filmsticking iteration from local minima by applying appropriate geometric and topological changes of the surface mesh. The algorithm is fully automatic and produces high-quality surface meshes for non-perfect inputs on both simulated and real scans.

   @article{wang2022restricted,
     title = {Restricted Delaunay Triangulation for Explicit Surface Reconstruction},
     author = {Wang, Pengfei and Wang, Zixiong and Xin, Shiqing and Gao, Xifeng and Wang, Wenping and Tu, Changhe},
     journal = {ACM Transactions on Graphics},
     year = {2022},
     publisher = {ACM},
     doi = {10.1145/3533768},
   }