Xiang Xu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Xiang Xu has a total of 27 co-authors. Here are the top co-authors by number of shared publications:
- Pradeep Kumar Jayaraman
School of Computer Science and Engineering, Nanyang Technological University, Singapore, Singapore
3 shared publications - Yilin Liu
University College London (UCL), London, United Kingdom
3 shared publications - Joseph George Lambourne
Autodesk AI Research Autodesk AI Research
2 shared publications - Baomin Xu
Beijing Jiaotong University
1 shared publication - Chinthala Reddy
Autodesk Research, London, United Kingdom
1 shared publication - Daniel Cohen-Or
Shenzhen University, Shenzhen, China
1 shared publication
Latest Publications
- Open Access
- 19 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-11
DualBrep: A Dual-Field Continuous Representation for B-rep Modelling
- 19 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-11
Open AccessDualBrep: A Dual-Field Continuous Representation for B-rep Modelling
- Open Access
- 14 December 2025
- Published by Association for Computing Machinery (ACM)
- p. 1-12
AutoBrep : Autoregressive B-Rep Generation with Unified Topology and Geometry
- 14 December 2025
- Published by Association for Computing Machinery (ACM)
- p. 1-12
Open AccessAutoBrep : Autoregressive B-Rep Generation with Unified Topology and Geometry
- Top Cited PapersOpen Access
- 27 July 2025
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Graphics
- Vol. 44 (4) , 1-25
HoLa: B-Rep Generation using a Holistic Latent Representation
We introduce a novel representation for learning and generating Computer-Aided Design (CAD) models in the form of boundary representations (B-Reps). Our representation unifies the continuous geometric properties of B-Rep primitives in different orders (e.g., surfaces and curves) and their discrete topological relations in a holistic latent (HoLa) space. This is based on the simple observation that the topological connection between two surfaces is intrinsically tied to the geometry of their intersecting curve. Such a prior allows us to reformulate topology learning in B-Reps as a geometric reconstruction problem in Euclidean space. Specifically, we eliminate the presence of curves, vertices, and all the topological connections in the latent space by learning to distinguish and derive curve geometries from a pair of surface primitives via a neural intersection network. To this end, our holistic latent space is only defined on surfaces but encodes a full B-Rep model, including the geometry of surfaces, curves, vertices, and their topological relations. Our compact and holistic latent space facilitates the design of a first diffusion-based generator to take on a large variety of inputs including point clouds, single/multi-view images, 2D sketches, and text prompts. Our method significantly reduces ambiguities, redundancies, and incoherences among the generated B-Rep primitives, as well as training complexities inherent in prior multi-step B-Rep learning pipelines, while achieving greatly improved validity rate over current state of the art: 82% vs. ≈50%.
- 27 July 2025
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Graphics
- Vol. 44 (4) , 1-25
Top Cited PapersOpen AccessHoLa: B-Rep Generation using a Holistic Latent Representation
We introduce a novel representation for learning and generating Computer-Aided Design (CAD) models in the form of boundary representations (B-Reps). Our representation unifies the continuous geometric properties of B-Rep primitives in different orders (e.g., surfaces and curves) and their discrete topological relations in a holistic latent (HoLa) space. This is based on the simple observation that the topological connection between two surfaces is intrinsically tied to the geometry of their intersecting curve. Such a prior allows us to reformulate topology learning in B-Reps as a geometric reconstruction problem in Euclidean space. Specifically, we eliminate the presence of curves, vertices, and all the topological connections in the latent space by learning to distinguish and derive curve geometries from a pair of surface primitives via a neural intersection network. To this end, our holistic latent space is only defined on surfaces but encodes a full B-Rep model, including the geometry of surfaces, curves, vertices, and their topological relations. Our compact and holistic latent space facilitates the design of a first diffusion-based generator to take on a large variety of inputs including point clouds, single/multi-view images, 2D sketches, and text prompts. Our method significantly reduces ambiguities, redundancies, and incoherences among the generated B-Rep primitives, as well as training complexities inherent in prior multi-step B-Rep learning pipelines, while achieving greatly improved validity rate over current state of the art: 82% vs. ≈50%.
- Top Cited PapersOpen Access
- 19 July 2024
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Graphics
- Vol. 43 (4) , 1-14
BrepGen: A B-rep Generative Diffusion Model with Structured Latent Geometry
This paper presents BrepGen , a diffusion-based generative approach that directly outputs a Boundary representation (B-rep) Computer-Aided Design (CAD) model. BrepGen represents a B-rep model as a novel structured latent geometry in a hierarchical tree. With the root node representing a whole CAD solid, each element of a B-rep model (i.e., a face, an edge, or a vertex) progressively turns into a child-node from top to bottom. B-rep geometry information goes into the nodes as the global bounding box of each primitive along with a latent code describing the local geometric shape. The B-rep topology information is implicitly represented by node duplication. When two faces share an edge, the edge curve will appear twice in the tree, and a T-junction vertex with three incident edges appears six times in the tree with identical node features. Starting from the root and progressing to the leaf, BrepGen employs Transformer-based diffusion models to sequentially denoise node features while duplicated nodes are detected and merged, recovering the B-Rep topology information. Extensive experiments show that BrepGen advances the task of CAD B-rep generation, surpassing existing methods on various benchmarks. Results on our newly collected furniture dataset further showcase its exceptional capability in generating complicated geometry. While previous methods were limited to generating simple prismatic shapes, BrepGen incorporates free-form and doubly-curved surfaces for the first time. Additional applications of BrepGen include CAD autocomplete and design interpolation. The code, pretrained models, and dataset are available at https://github.com/samxuxiang/BrepGen.
- 19 July 2024
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Graphics
- Vol. 43 (4) , 1-14
Top Cited PapersOpen AccessBrepGen: A B-rep Generative Diffusion Model with Structured Latent Geometry
This paper presents BrepGen , a diffusion-based generative approach that directly outputs a Boundary representation (B-rep) Computer-Aided Design (CAD) model. BrepGen represents a B-rep model as a novel structured latent geometry in a hierarchical tree. With the root node representing a whole CAD solid, each element of a B-rep model (i.e., a face, an edge, or a vertex) progressively turns into a child-node from top to bottom. B-rep geometry information goes into the nodes as the global bounding box of each primitive along with a latent code describing the local geometric shape. The B-rep topology information is implicitly represented by node duplication. When two faces share an edge, the edge curve will appear twice in the tree, and a T-junction vertex with three incident edges appears six times in the tree with identical node features. Starting from the root and progressing to the leaf, BrepGen employs Transformer-based diffusion models to sequentially denoise node features while duplicated nodes are detected and merged, recovering the B-Rep topology information. Extensive experiments show that BrepGen advances the task of CAD B-rep generation, surpassing existing methods on various benchmarks. Results on our newly collected furniture dataset further showcase its exceptional capability in generating complicated geometry. While previous methods were limited to generating simple prismatic shapes, BrepGen incorporates free-form and doubly-curved surfaces for the first time. Additional applications of BrepGen include CAD autocomplete and design interpolation. The code, pretrained models, and dataset are available at https://github.com/samxuxiang/BrepGen.
- Open Access
Hierarchical Neural Coding for Controllable CAD Model Generation
This paper presents a novel generative model for Computer Aided Design (CAD) that 1) represents high-level design concepts of a CAD model as a three-level hierarchical tree of neural codes, from global part arrangement down to local curve geometry; and 2) controls the generation or completion of CAD models by specifying the target design using a code tree. Concretely, a novel variant of a vector quantized VAE with "masked skip connection" extracts design variations as neural codebooks at three levels. Two-stage cascaded auto-regressive transformers learn to generate code trees from incomplete CAD models and then complete CAD models following the intended design. Extensive experiments demonstrate superior performance on conventional tasks such as random generation while enabling novel interaction capabilities on conditional generation tasks. The code is available at https://github.com/samxuxiang/hnc-cad.
Open AccessHierarchical Neural Coding for Controllable CAD Model Generation
This paper presents a novel generative model for Computer Aided Design (CAD) that 1) represents high-level design concepts of a CAD model as a three-level hierarchical tree of neural codes, from global part arrangement down to local curve geometry; and 2) controls the generation or completion of CAD models by specifying the target design using a code tree. Concretely, a novel variant of a vector quantized VAE with "masked skip connection" extracts design variations as neural codebooks at three levels. Two-stage cascaded auto-regressive transformers learn to generate code trees from incomplete CAD models and then complete CAD models following the intended design. Extensive experiments demonstrate superior performance on conventional tasks such as random generation while enabling novel interaction capabilities on conditional generation tasks. The code is available at https://github.com/samxuxiang/hnc-cad.
- 25 June 2022
- Published by Springer Nature in Applied Intelligence
- Vol. 53 (5) , 1-13
An enhancement model based on dense atrous and inception convolution for image semantic segmentation
- 25 June 2022
- Published by Springer Nature in Applied Intelligence
- Vol. 53 (5) , 1-13
An enhancement model based on dense atrous and inception convolution for image semantic segmentation
- Open Access
MCMI: Multi-Cycle Image Translation with Mutual Information Constraints
We present a mutual information-based framework for unsupervised image-to-image translation. Our MCMI approach treats single-cycle image translation models as modules that can be used recurrently in a multi-cycle translation setting where the translation process is bounded by mutual information constraints between the input and output images. The proposed mutual information constraints can improve cross-domain mappings by optimizing out translation functions that fail to satisfy the Markov property during image translations. We show that models trained with MCMI produce higher quality images and learn more semantically-relevant mappings compared to state-of-the-art image translation methods. The MCMI framework can be applied to existing unpaired image-to-image translation models with minimum modifications. Qualitative experiments and a perceptual study demonstrate the image quality improvements and generality of our approach using several backbone models and a variety of image datasets.
Open AccessMCMI: Multi-Cycle Image Translation with Mutual Information Constraints
We present a mutual information-based framework for unsupervised image-to-image translation. Our MCMI approach treats single-cycle image translation models as modules that can be used recurrently in a multi-cycle translation setting where the translation process is bounded by mutual information constraints between the input and output images. The proposed mutual information constraints can improve cross-domain mappings by optimizing out translation functions that fail to satisfy the Markov property during image translations. We show that models trained with MCMI produce higher quality images and learn more semantically-relevant mappings compared to state-of-the-art image translation methods. The MCMI framework can be applied to existing unpaired image-to-image translation models with minimum modifications. Qualitative experiments and a perceptual study demonstrate the image quality improvements and generality of our approach using several backbone models and a variety of image datasets.
- 1 November 2013
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 24, 93-98
Face hallucination: How much it can improve face recognition
Face hallucination has been a popular topic in image processing in recent years. Currently the commonly used performance criteria for face hallucination are peak signal noise ratio (PSNR) and the root mean square error (RMSE). Though it is logically believed that hallucinated high-resolution face images should have a better performance in face recognition, we show in this paper that this `the higher resolution, the higher recognition' assumption is not validated systematically by some designed experiments. First, we illustrate this assumption only works when the image solution is sufficiently large. Second, in the case of very extreme low resolutions, the recognition performance of the hallucinated images obtained by some typical existing face hallucination approaches will not improve. Finally, the relationship of the popular evaluation methods in face hallucination, PSNR and RMSE, with the recognition performance are investigated. The findings of this paper can help people design new hallucination approaches with an aim of improving face recognition performance with specified classifiers.
- 1 November 2013
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 24, 93-98
Face hallucination: How much it can improve face recognition
Face hallucination has been a popular topic in image processing in recent years. Currently the commonly used performance criteria for face hallucination are peak signal noise ratio (PSNR) and the root mean square error (RMSE). Though it is logically believed that hallucinated high-resolution face images should have a better performance in face recognition, we show in this paper that this `the higher resolution, the higher recognition' assumption is not validated systematically by some designed experiments. First, we illustrate this assumption only works when the image solution is sufficiently large. Second, in the case of very extreme low resolutions, the recognition performance of the hallucinated images obtained by some typical existing face hallucination approaches will not improve. Finally, the relationship of the popular evaluation methods in face hallucination, PSNR and RMSE, with the recognition performance are investigated. The findings of this paper can help people design new hallucination approaches with an aim of improving face recognition performance with specified classifiers.