Pith. sign in

REVIEW 4 major objections 6 minor 52 references

CMT: A Cascade MAR with Topology Predictor for Multimodal Conditional CAD Generation

T0 review · 4 major / 6 minor · reviewed 2026-08-16 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read CMT claims that generating CAD edges first and surfaces second, with a learned topology predictor, produces more accurate and more valid solid models from text, point clouds, or images.

desk verdict First multimodal B-Rep CAD generation with a strong dataset, but the empirical claims are weakened by missing error bars, a mismatched baseline, and an unablated topology predictor. read the letter →

arxiv 2504.20830 v2 pith:X7X4JSQN submitted 2025-04-29 cs.CV

Jianyu Wu , Yizhou Wang , Xiangyu Yue , Xinzhu Ma , Jingyang Guo , Dongzhan Zhou , Wanli Ouyang , Shixiang Tang This is my paper · ORCID
classification cs.CV
keywords CADgenerationBoundaryRepresentationB-Repmaskedautoregressivemodelmultimodalconditioningtopologypredictionpointcloudimage-conditioned
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

CMT is a proposed framework for generating Boundary Representation (B-Rep) CAD models from multiple user input types—text, point clouds, and multi-view images—in a single autoregressive model. Its central claim is that generating a model in two cascade stages, edges first and then surfaces, with a small learned module that predicts which edges bound which surfaces, produces CAD geometry that is closer to real designs and more often topologically valid than previous command-sequence or one-shot diffusion generators. To train this, the authors assemble mmABC, a dataset of 1.35 million B-Rep models with captions, point clouds, and rendered images. If the approach holds, CAD automation shifts from reconstructing simplified command histories to directly producing manufacturable solid models from everyday inputs.

What carries the argument

The load-bearing mechanism is the cascade masked autoregressive network: an Edge MAR generates the ordered edge tokens first, then a Surface MAR generates surface tokens conditioned on both the user-input embedding and a fixed-length summary of the generated edges, following the B-Rep fact that edges contour surfaces. Tokens are continuous VAE codes augmented with bounding-box coordinates, decoded by a diffusion MLP. A cross-attention topology predictor then produces an edge-surface adjacency matrix, thresholded at 0.5, to assemble the final B-Rep.

What would settle it

Sweep the topology-predictor threshold on ABC and watch Valid ratio: if validity does not peak near the default 0.5, or if a simple nearest-edge-to-surface matching rule achieves the same validity, the paper's topology-prediction component is not what drives the reported validity gains.

Watch

Extended reading notes

Core claim

On its own terms, the paper's discovery is that B-Rep generation can be decomposed as an edges-then-surfaces autoregressive diffusion process, and that this ordering alone embeds the B-Rep prior that faces are bounded by edges. The paper reports that on the ABC dataset, unconditional generation improves Coverage by +10.68 percentage points and Valid ratio by +10.3 points over BrepGen, the previous B-Rep diffusion baseline; on DeepCAD, Valid rises to 70.1%. For image-conditioned generation, Chamfer distance drops from 6.18 (InstantMesh) to 2.17, and for point-cloud conditioning the method exceeds a dedicated reconstructor on F-score (99.07 vs 98.17) and Normal Consistency (95.48 vs 94.36). The same recipe handles text, point cloud, and image conditioning through one frozen vision encoder plus a learned projector.

Load-bearing premise

The load-bearing assumption is that the learned cross-attention topology predictor, with its fixed 0.5 cutoff, recovers the correct edge-surface adjacency graph; if that matrix is wrong, the assembled model is not a valid B-Rep even when every generated edge and surface is geometrically perfect.

Editorial extensions

If this is right

  • A single autoregressive model can serve both unconditional generation and conditioning on text, images, or point clouds, removing the need for separate pipelines per modality.
  • Topology validity becomes a training signal rather than a post-processing chore: the cascade and topology predictor together raise the fraction of quantized outputs that survive as valid B-Reps, to 70.1% on DeepCAD and 58.5% on ABC.
  • Image-conditioned B-Rep generation now approaches the accuracy of general single-image mesh reconstruction while retaining CAD-native surfaces and edges.
  • The mmABC dataset gives the community a 1.35-million-model benchmark with paired text, point cloud, and image annotations, making conditional B-Rep generation trainable at scale.
  • Autoregressive sampling depth can be traded against quality: reducing generation steps from full sequence length to 1/1 collapses Valid ratio to 0.10%, showing that step-by-step generation is essential.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The fixed 0.5 threshold on the adjacency matrix is an obvious point of attack: sweeping or learning the threshold could recover a meaningful share of the roughly 40% of ABC outputs still counted invalid, since topology prediction is the last assembly step.
  • Because the reported speedup over the Point2CAD post-processor is more than 4200x, the topology predictor could also be reused as a validation or repair module on top of any B-Rep token generator, not only this one.
  • The cascade's value likely transfers to other structured outputs with a similar containment hierarchy, such as sketch-and-extrude design histories, though the paper does not test that.
  • mmABC's text captions are generated by a vision-language model rather than written by humans, so downstream text-conditioned quality may partly reflect caption diversity; a human-verified subset would be a useful follow-up.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. The paper introduces CMT, a cascade masked autoregressive framework for B-Rep CAD generation conditioned on text, images, and point clouds. It tokenizes surfaces and edges as continuous tokens, generates edges before surfaces using two MARs, and adds a cross-attention topology predictor that recovers the edge-face adjacency matrix. The authors also construct mmABC, a 1.35-million-model multimodal B-Rep dataset built from ABC, with rendered images, point clouds, and VLM-generated captions. Experiments report improvements over BrepGen on unconditional DeepCAD and ABC generation, and over several reconstruction and generation baselines on point-cloud, image, and text-conditioned tasks.

Significance. If the results hold, CMT would be a notable step in CAD generation: it is the first multimodal B-Rep generator, and the mmABC dataset is a substantial community resource. The cascade edge-to-surface design is well motivated by the structure of B-Reps, and the reported topology-prediction speedup is practically useful. However, the empirical support for the central claims is weakened by a contradictory metric outcome in Table 2, the absence of uncertainty estimates, and the lack of an ablation isolating the topology predictor. The architecture-level idea remains defensible, so the paper is worth revising rather than rejecting.

major comments (4)
  1. [§4.4, Eq. (3), Table 2] The topology predictor is the load-bearing component for the final B-Rep assembly, but the paper provides no evidence that it enforces B-Rep validity. The objective in Eq. (3) is an elementwise MSE on adjacency scores followed by a fixed threshold τ=0.5; no term encodes global B-Rep constraints, such as every edge being incident to exactly two faces, closed face boundaries, or absence of face intersections. Table 2 shows the consequence: Valid is 70.1% on DeepCAD and 58.5% on ABC, meaning roughly 30–42% of generated models fail at assembly. Table 6 ablates the cascade and sampling steps but keeps the topology predictor fixed, so the reported +10.3% Valid gain over BrepGen cannot be attributed to the predictor. Please add an ablation that varies or removes the topology predictor, report precision/recall of adjacency prediction, and discuss how invalid topologies are handled.
  2. [§5.2, Table 2] The text claims "consistent and remarkable improvements on COV, MMD and JSD," but Table 2 shows JSD on DeepCAD increases from 0.09 (BrepGen) to 1.02 (CMT), an order-of-magnitude degradation. The ABC row also ties MMD at 1.35. Please correct the claim, discuss the diversity/quality trade-off, and report token-level or distribution-level diagnostics that explain the JSD gap.
  3. [§5.1, Tables 2–4] All quantitative results are single-seed point estimates with no error bars. Because sampling from an autoregressive diffusion model is stochastic and several headline differences are small (MMD 1.35 vs 1.35 on ABC, and F-score margins of less than one point in Table 3), please report means and standard deviations over at least three seeds, or bootstrap confidence intervals, before claiming state-of-the-art accuracy.
  4. [§5.3, Tables 3–5] The conditional comparisons do not isolate the contribution of the proposed method. InstantMesh is a general mesh generator, not a B-Rep or CAD-command baseline, and DeepCAD is a command-sequence model that may be trained on a different dataset; the text-conditioned evaluation uses only 20 models and reports win rates without confidence intervals. Please compare against B-Rep-capable baselines on the same data splits and report the evaluation protocol in full.
minor comments (6)
  1. [Abstract and Introduction] The manuscript contains typos such as "c ascade", "conditonal", and "mulitview"; please copyedit the text.
  2. [Eq. (1)–(2)] The notation in the diffusion losses is malformed: "e_{t,|t}" and "s_{t,|t}" should presumably be "e_t" and "s_t" with the noise level t, and the norm expressions should be typeset as squared L2 norms.
  3. [Figure 2] The caption says the topology predictor operates on "generated surface tokens ˆE", but this should be "ˆS" to match the notation in the text.
  4. [§3] The VLM-generated text annotations are produced with random prompts from InternVL2-40B, but no human agreement or filtering statistics are reported; please add quality checks for the captions, since noisy text directly affects the text-conditioned evaluation.
  5. [§5.3, Table 5] The text-conditioned win rate is based on only 20 models; please include confidence intervals, a larger sample, or both, and specify how the 20 models were selected.
  6. [§5.1] The construction of the mmABC split should be stated more precisely: the test set includes all ABC test models plus 10% of the extra decoupled models, which complicates direct comparability with prior ABC results.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: core generations are trained and evaluated on external corpora and baselines; no fitted quantity is renamed as a prediction.

full rationale

I find no circularity in the paper's derivation chain. The method's central components — continuous tokenization (Sec. 4.1), the cascade autoregressive network (Sec. 4.3), and the topology predictor (Sec. 4.4) — are trained with the explicit losses Ledge, Lsurf, Ltopo in Eqs. (1)-(4) on the external ABC and DeepCAD corpora for unconditional generation. The reported gains (Table 2) are measured against external baselines (BrepGen, SolidGen, DeepCAD) using geometric distribution metrics, and the Valid ratio is determined by geometric validity of the assembled B-Rep after quantization, not by the topology predictor's own adjacency scores. Conditional generation is evaluated on the newly constructed mmABC against external reconstruction and generation methods (Point2CAD, NVDNet, InstantMesh, Michelangelo) and by external judges (GPT-4o, Qwen-2.5-VL, human CAD designers). No constant is fitted to a subset of the evaluation data and then presented as a prediction, and no load-bearing argument reduces to a self-citation. The few self-citations ([12], [27]) are background references unrelated to the method's core claims. The fixed 0.5 threshold on the topology predictor is a robustness/correctness concern, not a circularity, because the validity metric is externally defined and the reported Valid ratios show the predictor failing on a substantial fraction of generations.

Assumptions & free parameters 4 free parameters · 4 assumptions · 0 invented entities

The central contributions are architectural and data-driven; the free parameters are the key manually-set hyperparameters. The main assumed priors are B-Rep-specific: the edge-to-surface cascade, the ordering heuristic, the sufficiency of the topology predictor, and VAE fidelity. No new physical entities or forces are introduced.

free parameters (4)
  • topology threshold tau = 0.5
    Chosen by hand in Sec 4.4 to binarize adjacency scores.
  • max edge sequence length = 128 (ABC/mmABC), 64 (DeepCAD)
    Chosen to cover over 95% of data per dataset (Sec 5.1).
  • max surface sequence length = 64 (ABC/mmABC), 32 (DeepCAD)
    Chosen to cover over 95% of data per dataset (Sec 5.1).
  • condition embedding size = R18x768
    Set in Sec 5.1 as the dimension of the fixed-length condition embedding.
assumptions (4)
  • domain assumption B-Rep generation can be decomposed as edges-then-surfaces (edge-contours-surface prior)
    Sec 4.3; the cascade MAR relies on this prior. If surfaces cannot be predicted from edge tokens alone, the decomposition fails. The ablation shows benefit, but it is an assumption about B-Rep structure.
  • domain assumption The bounding-box-based token ordering yields a stable canonical sequence for autoregressive modeling
    Sec 4.1, token ordering. Tokens are sorted lexicographically by bounding-box coordinates; if this ordering is not stable under rotation or minor geometry changes, conditional dependencies may be misaligned.
  • domain assumption A thresholded cross-attention layer can predict the full edge-surface adjacency graph
    Sec 4.4; the topology predictor uses one cross-attention layer and threshold 0.5 to decide adjacency. No structural constraints, such as manifoldness, are enforced.
  • domain assumption The VAE tokenizers reconstruct edge and surface geometry accurately enough for valid B-Rep assembly
    Sec 4.1; lossy VAE encoding followed by denoising must preserve geometric and topological consistency; no quantization error analysis is provided.

how reviews work

0 comments
Cite this review

Pith. "Pith review of CMT: A Cascade MAR with Topology Predictor for Multimodal Conditional CAD Generation." pith.science (2026). https://pith.science/paper/X7X4JSQN

@misc{pith2026250420830,
  author       = {Pith},
  title        = {Pith review of: CMT: A Cascade MAR with Topology Predictor for Multimodal Conditional CAD Generation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/X7X4JSQN}},
  note         = {Machine review of arXiv:2504.20830}
}
read the original abstract

While accurate and user-friendly Computer-Aided Design (CAD) is crucial for industrial design and manufacturing, existing methods still struggle to achieve this due to their over-simplified representations or architectures incapable of supporting multimodal design requirements. In this paper, we attempt to tackle this problem from both methods and datasets aspects. First, we propose a cascade MAR with topology predictor (CMT), the first multimodal framework for CAD generation based on Boundary Representation (B-Rep). Specifically, the cascade MAR can effectively capture the ``edge-counters-surface'' priors that are essential in B-Reps, while the topology predictor directly estimates topology in B-Reps from the compact tokens in MAR. Second, to facilitate large-scale training, we develop a large-scale multimodal CAD dataset, mmABC, which includes over 1.3 million B-Rep models with multimodal annotations, including point clouds, text descriptions, and multi-view images. Extensive experiments show the superior of CMT in both conditional and unconditional CAD generation tasks. For example, we improve Coverage and Valid ratio by +10.68% and +10.3%, respectively, compared to state-of-the-art methods on ABC in unconditional generation. CMT also improves +4.01 Chamfer on image conditioned CAD generation on mmABC.

Figures

Figures reproduced from arXiv: 2504.20830 by the authors.

Figure 1
Figure 1. CMT generates accurate and diverse CAD models of Boundary Represention based on user’s multimodal inputs, including texts, point clouds and mulitview images, which can be seamless applied in the design and industrial manufacturing. Abstract While accurate and user-friendly Computer-Aided Design (CAD) is crucial for industrial design and manufacturing, existing methods still struggle to achieve this due to their over… view at source ↗
Figure 2
Figure 2. The pipeline of CMT. CMT consists of continuous tokenization of B-Rep, a unified multimodal condition encoder, a cascade autoregressive generation network, and a topology predictor. Given a CAD model B and its corresponding multimodal inputs C, (a) the continuous tokenization process in CMT converts the B-Rep into continuous surface tokens S and edge tokens E using Surface VAE [20] and Edge VAE respectively. (b) the… view at source ↗
Figure 3
Figure 3. Unconditional generation results on ABC [ [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (1 more)
Figure 5
Figure 5. Figure 5: Comparisons on image-conditional generation task. [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

52 extracted references · 33 canonical work pages

  1. [1]

    Gencad: Image- conditioned computer-aided design generation with transformer-based contrastive representation and diffusion priors

    Md Ferdous Alam and Faez Ahmed. Gencad: Image- conditioned computer-aided design generation with transformer-based contrastive representation and diffusion priors. arXiv preprint arXiv:2409.16294, 2024. 2, 3

    arXiv 2024

  2. [2]

    Query2cad: Generating cad models using natural language queries

    Akshay Badagabettu, Sai Sravan Yarlagadda, and Amir Barati Farimani. Query2cad: Generating cad models using natural language queries. arXiv preprint arXiv:2406.00144, 2024. 2

    arXiv 2024

  3. [3]

    Shuai Bai, Keqin Chen, Xuejing Liu, Jialin Wang, Wenbin Ge, Sibo Song, Kai Dang, Peng Wang, Shijie Wang, Jun Tang, et al. Qwen2. 5-vl technical report. arXiv preprint arXiv:2502.13923, 2025. 8

    arXiv 2025

  4. [4]

    Algorithmic differentiation of the open cascade technology cad kernel and its coupling with an adjoint cfd solver

    Mladen Banovi ´c, Orest Mykhaskiv, Salvatore Auriemma, Andrea Walther, Herve Legrand, and Jens-Dominik M ¨uller. Algorithmic differentiation of the open cascade technology cad kernel and its coupling with an adjoint cfd solver. Opti- mization Methods and Software, 33(4-6):813–828, 2018. 4

    work page 2018

  5. [5]

    Computer aided detection (cad): an overview

    Ronald A Castellino. Computer aided detection (cad): an overview. Cancer Imaging, 5(1):17, 2005. 2

    work page 2005

  6. [6]

    Meshxl: Neural coordinate field for generative 3d foundation models

    Sijin Chen, Xin Chen, Anqi Pang, Xianfang Zeng, Wei Cheng, Yijun Fu, Fukun Yin, Billzb Wang, Jingyi Yu, Gang Yu, et al. Meshxl: Neural coordinate field for generative 3d foundation models. Advances in Neural Information Pro- cessing Systems, 37:97141–97166, 2024. 5

    work page 2024

  7. [7]

    Img2cad: Conditioned 3d cad model generation from single image with structured visual geometry

    Tianrun Chen, Chunan Yu, Yuanqi Hu, Jing Li, Tao Xu, Run- long Cao, Lanyun Zhu, Ying Zang, Yong Zhang, Zejian Li, et al. Img2cad: Conditioned 3d cad model generation from single image with structured visual geometry. arXiv preprint arXiv:2410.03417, 2024. 2, 3

    arXiv

  8. [8]

    Internvl: Scaling up vision foundation mod- els and aligning for generic visual-linguistic tasks

    Zhe Chen, Jiannan Wu, Wenhai Wang, Weijie Su, Guo Chen, Sen Xing, Muyan Zhong, Qinglong Zhang, Xizhou Zhu, Lewei Lu, et al. Internvl: Scaling up vision foundation mod- els and aligning for generic visual-linguistic tasks. In Pro- ceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 24185–24198, 2024. 4

    work page 2024

Show all 52 references
  1. [9]

    Cadops-net: Jointly learning cad operation types and steps from boundary-representations

    Elona Dupont, Kseniya Cherenkova, Anis Kacem, Sk Aziz Ali, Ilya Arzhannikov, Gleb Gusev, and Djamila Aouada. Cadops-net: Jointly learning cad operation types and steps from boundary-representations. In 2022 International Con- ference on 3D Vision (3DV), pages 114–123. IEEE, 2022. 3

    2022

  2. [10]

    Transcad: A hi- erarchical transformer for cad sequence inference from point clouds

    Elona Dupont, Kseniya Cherenkova, Dimitrios Mallis, Gleb Gusev, Anis Kacem, and Djamila Aouada. Transcad: A hi- erarchical transformer for cad sequence inference from point clouds. In European Conference on Computer Vision, pages 19–36. Springer, 2024. 2

    2024

  3. [11]

    Complexgen: Cad reconstruction by b-rep chain complex generation

    Haoxiang Guo, Shilin Liu, Hao Pan, Yang Liu, Xin Tong, and Baining Guo. Complexgen: Cad reconstruction by b-rep chain complex generation. ACM Transactions on Graphics (TOG), 41(4):1–18, 2022. 7

    2022

  4. [12]

    Compressing large language models by joint spar- sification and quantization

    Jinyang Guo, Jianyu Wu, Zining Wang, Jiaheng Liu, Ge Yang, Yifu Ding, Ruihao Gong, Haotong Qin, and Xiang- long Liu. Compressing large language models by joint spar- sification and quantization. In Forty-first International Con- ference on Machine Learning, 2024. 3

    2024

  5. [13]

    Onellm: One framework to align all modalities with language

    Jiaming Han, Kaixiong Gong, Yiyuan Zhang, Jiaqi Wang, Kaipeng Zhang, Dahua Lin, Yu Qiao, Peng Gao, and Xi- angyu Yue. Onellm: One framework to align all modalities with language. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 26584–...

    2024

  6. [14]

    Deep residual learning for image recognition

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recognition. In Proceed- ings of the IEEE conference on computer vision and pattern recognition, pages 770–778, 2016. 2

    2016

  7. [15]

    Denoising dif- fusion probabilistic models

    Jonathan Ho, Ajay Jain, and Pieter Abbeel. Denoising dif- fusion probabilistic models. Advances in neural information processing systems, 33:6840–6851, 2020. 2, 3, 5, 6

    2020

  8. [16]

    Gpt-4o system card

    Aaron Hurst, Adam Lerer, Adam P Goucher, Adam Perel- man, Aditya Ramesh, Aidan Clark, AJ Ostrow, Akila Weli- hinda, Alan Hayes, Alec Radford, et al. Gpt-4o system card. arXiv preprint arXiv:2410.21276, 2024. 8

    2024 arXiv

  9. [17]

    Solidgen: An autoregressive model for direct b-rep synthesis

    Pradeep Kumar Jayaraman, Joseph George Lambourne, Nishkrit Desai, Karl Willis, Aditya Sanghi, and Nigel JW Morris. Solidgen: An autoregressive model for direct b-rep synthesis. Transactions on Machine Learning Research. 2, 3, 6

  10. [18]

    Cad-signet: Cad language inference from point clouds using layer-wise sketch instance guided attention

    Mohammad Sadil Khan, Elona Dupont, Sk Aziz Ali, Kseniya Cherenkova, Anis Kacem, and Djamila Aouada. Cad-signet: Cad language inference from point clouds using layer-wise sketch instance guided attention. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pat- ter...

    2024

  11. [19]

    Text2cad: Generating sequential cad designs from beginner- to-expert level text prompts.Advances in Neural Information Processing Systems, 37:7552–7579, 2024

    Mohammad Sadil Khan, Sankalp Sinha, Talha Uddin, Di- dier Stricker, Sk Aziz Ali, and Muhammad Zeshan Afzal. Text2cad: Generating sequential cad designs from beginner- to-expert level text prompts.Advances in Neural Information Processing Systems, 37:7552–7579, 2024. 2, 3

    2024

  12. [20]

    Auto-encoding vari- ational bayes

    Diederik P Kingma, Max Welling, et al. Auto-encoding vari- ational bayes. 4, 5

  13. [21]

    Abc: A big cad model dataset for geometric deep learning

    Sebastian Koch, Albert Matveev, Zhongshi Jiang, Francis Williams, Alexey Artemov, Evgeny Burnaev, Marc Alexa, Denis Zorin, and Daniele Panozzo. Abc: A big cad model dataset for geometric deep learning. In Proceedings of the IEEE/CVF conference on computer vision and pattern re...

    2019

  14. [22]

    Brepnet: A topological message passing system for solid models

    Joseph G Lambourne, Karl DD Willis, Pradeep Kumar Jayaraman, Aditya Sanghi, Peter Meltzer, and Hooman Shayani. Brepnet: A topological message passing system for solid models. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages 12773– 12...

    2021

  15. [23]

    Free2cad: Parsing freehand drawings into cad commands

    Changjian Li, Hao Pan, Adrien Bousseau, and Niloy J Mitra. Free2cad: Parsing freehand drawings into cad commands. ACM Transactions on Graphics (TOG), 41(4):1–16, 2022. 3

    2022

  16. [24]

    Blip-2: Bootstrapping language-image pre-training with frozen image encoders and large language models

    Junnan Li, Dongxu Li, Silvio Savarese, and Steven Hoi. Blip-2: Bootstrapping language-image pre-training with frozen image encoders and large language models. In In- ternational conference on machine learning , pages 19730– 19742. PMLR, 2023. 3

    2023

  17. [25]

    Autoregressive image generation without vec- tor quantization

    Tianhong Li, Yonglong Tian, He Li, Mingyang Deng, and Kaiming He. Autoregressive image generation without vec- tor quantization. Advances in Neural Information Processing Systems, 37:56424–56445, 2024. 1, 2, 5, 6

    2024

  18. [26]

    Visual instruction tuning

    Haotian Liu, Chunyuan Li, Qingyang Wu, and Yong Jae Lee. Visual instruction tuning. Advances in neural information processing systems, 36:34892–34916, 2023. 3

    2023

  19. [27]

    Afbench: A large-scale benchmark for airfoil design

    Jian Liu, Jianyu Wu, Hairun Xie, Jing Wang, Liu Wei, Wanli Ouyang, Junjun Jiang, Xianming Liu, SHIXIANG TANG, Miao Zhang, et al. Afbench: A large-scale benchmark for airfoil design. Advances in Neural Information Processing Systems, 37:82757–82780, 2025. 2

    2025

  20. [28]

    Split-and-fit: Learning b-reps via structure-aware voronoi partitioning

    Yilin Liu, Jiale Chen, Shanshan Pan, Daniel Cohen-Or, Hao Zhang, and Hui Huang. Split-and-fit: Learning b-reps via structure-aware voronoi partitioning. ACM Transactions on Graphics (TOG), 43(4):1–13, 2024. 7

    2024

  21. [29]

    Point2cad: Reverse engineering cad models from 3d point clouds

    Yujia Liu, Anton Obukhov, Jan Dirk Wegner, and Konrad Schindler. Point2cad: Reverse engineering cad models from 3d point clouds. In Proceedings of the IEEE/CVF Confer- ence on Computer Vision and Pattern Recognition , pages 3763–3772, 2024. 2, 7

    2024

  22. [30]

    Learning transferable visual models from natural language supervi- sion

    Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, et al. Learning transferable visual models from natural language supervi- sion. In International conference on machine learning, ...

    2021

  23. [31]

    Mlcad: A survey of research in machine learning for cad keynote paper

    Martin Rapp, Hussam Amrouch, Yibo Lin, Bei Yu, David Z Pan, Marilyn Wolf, and J ¨org Henkel. Mlcad: A survey of research in machine learning for cad keynote paper. IEEE Transactions on Computer-Aided Design of Integrated Cir- cuits and Systems, 41(10):3162–3181, 2021. 2

    2021

  24. [32]

    Gener- ating diverse high-fidelity images with vq-vae-2

    Ali Razavi, Aaron Van den Oord, and Oriol Vinyals. Gener- ating diverse high-fidelity images with vq-vae-2. Advances in neural information processing systems, 32, 2019. 3

    2019

  25. [33]

    Faster r-cnn: Towards real-time object detection with region proposal networks

    Shaoqing Ren, Kaiming He, Ross Girshick, and Jian Sun. Faster r-cnn: Towards real-time object detection with region proposal networks. IEEE transactions on pattern analysis and machine intelligence, 39(6):1137–1149, 2016. 2

    2016

  26. [34]

    Sketchgraphs: A large-scale dataset for modeling rela- tional geometry in computer-aided design

    Ari Seff, Yaniv Ovadia, Wenda Zhou, and Ryan P Adams. Sketchgraphs: A large-scale dataset for modeling rela- tional geometry in computer-aided design. arXiv preprint arXiv:2007.08506, 2020. 3

    2007 arXiv

  27. [35]

    Meshgpt: Generating triangle meshes with decoder-only transformers

    Yawar Siddiqui, Antonio Alliegro, Alexey Artemov, Ta- tiana Tommasi, Daniele Sirigatti, Vladislav Rosov, Angela Dai, and Matthias Nießner. Meshgpt: Generating triangle meshes with decoder-only transformers. In Proceedings of the IEEE/CVF conference on computer vision and patte...

    2024

  28. [36]

    Adapting object detec- tors with conditional domain normalization

    Peng Su, Kun Wang, Xingyu Zeng, Shixiang Tang, Dapeng Chen, Di Qiu, and Xiaogang Wang. Adapting object detec- tors with conditional domain normalization. In Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part XI 16 , pages 40...

    2020

  29. [37]

    Point2cyl: Reverse engineering 3d objects from point clouds to extrusion cylinders

    Mikaela Angelina Uy, Yen-Yu Chang, Minhyuk Sung, Purvi Goel, Joseph G Lambourne, Tolga Birdal, and Leonidas J Guibas. Point2cyl: Reverse engineering 3d objects from point clouds to extrusion cylinders. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Re...

    2022

  30. [38]

    Neural discrete representation learning

    Aaron Van Den Oord, Oriol Vinyals, et al. Neural discrete representation learning. Advances in neural information pro- cessing systems, 30, 2017. 3

    2017

  31. [39]

    Attention is all you need

    Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszko- reit, Llion Jones, Aidan N Gomez, Łukasz Kaiser, and Illia Polosukhin. Attention is all you need. Advances in neural information processing systems, 30, 2017. 2, 4, 5

    2017

  32. [40]

    Fusion 360 gallery: A dataset and environ- ment for programmatic cad construction from human design sequences

    Karl DD Willis, Yewen Pu, Jieliang Luo, Hang Chu, Tao Du, Joseph G Lambourne, Armando Solar-Lezama, and Wo- jciech Matusik. Fusion 360 gallery: A dataset and environ- ment for programmatic cad construction from human design sequences. ACM Transactions on Graphics (TOG) , 40(4)...

    2021

  33. [41]

    Deepcad: A deep generative network for computer-aided design models

    Rundi Wu, Chang Xiao, and Changxi Zheng. Deepcad: A deep generative network for computer-aided design models. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 6772–6782, 2021. 2, 3, 6, 7, 8

    2021

  34. [42]

    Next-gpt: Any-to-any multimodal llm

    Shengqiong Wu, Hao Fei, Leigang Qu, Wei Ji, and Tat-Seng Chua. Next-gpt: Any-to-any multimodal llm. In Forty-first International Conference on Machine Learning, 2024. 3

    2024

  35. [43]

    Instantmesh: Efficient 3d mesh generation from a single image with sparse-view large reconstruction models

    Jiale Xu, Weihao Cheng, Yiming Gao, Xintao Wang, Shenghua Gao, and Ying Shan. Instantmesh: Efficient 3d mesh generation from a single image with sparse-view large reconstruction models. arXiv preprint arXiv:2404.07191 ,

    arXiv

  36. [44]

    Cad-mllm: Unifying multimodality- conditioned cad generation with mllm

    Jingwei Xu, Chenyu Wang, Zibo Zhao, Wen Liu, Yi Ma, and Shenghua Gao. Cad-mllm: Unifying multimodality- conditioned cad generation with mllm. arXiv preprint arXiv:2411.04954, 2024. 2, 3, 6

    2024 arXiv

  37. [45]

    Skexgen: Autoregressive generation of cad construction se- quences with disentangled codebooks

    Xiang Xu, Karl DD Willis, Joseph G Lambourne, Chin-Yi Cheng, Pradeep Kumar Jayaraman, and Yasutaka Furukawa. Skexgen: Autoregressive generation of cad construction se- quences with disentangled codebooks. In International Con- ference on Machine Learning, pages 24698–24724. PMLR,

  38. [46]

    Hierar- chical neural coding for controllable cad model generation

    Xiang Xu, Pradeep Kumar Jayaraman, Joseph George Lam- bourne, Karl DD Willis, and Yasutaka Furukawa. Hierar- chical neural coding for controllable cad model generation. In International Conference on Machine Learning , pages 38443–38461. PMLR, 2023. 2, 3

    2023

  39. [47]

    Brepgen: A b-rep generative diffusion model with structured latent geometry

    Xiang Xu, Joseph Lambourne, Pradeep Jayaraman, Zhengqing Wang, Karl Willis, and Yasutaka Furukawa. Brepgen: A b-rep generative diffusion model with structured latent geometry. ACM Transactions on Graphics (TOG), 43 (4):1–14, 2024. 2, 3, 5, 6, 7

    2024

  40. [48]

    Img2cad: Reverse engineering 3d cad models from images through vlm-assisted conditional factorization

    Yang You, Mikaela Angelina Uy, Jiaqi Han, Rahul Thomas, Haotong Zhang, Suya You, and Leonidas Guibas. Img2cad: Reverse engineering 3d cad models from images through vlm-assisted conditional factorization. arXiv preprint arXiv:2408.01437, 2024. 3

    2024

  41. [49]

    Openecad: An efficient visual language model for editable 3d-cad design

    Zhe Yuan, Jianqi Shi, and Yanhong Huang. Openecad: An efficient visual language model for editable 3d-cad design. Computers & Graphics, 124:104048, 2024. 3

    2024

  42. [50]

    Michelangelo: Conditional 3d shape generation based on shape-image-text aligned latent representation

    Zibo Zhao, Wen Liu, Xin Chen, Xianfang Zeng, Rui Wang, Pei Cheng, BIN FU, Tao Chen, Gang YU, and Shenghua Gao. Michelangelo: Conditional 3d shape generation based on shape-image-text aligned latent representation. In Thirty- seventh Conference on Neural Information Processing ...

    2023

  43. [51]

    Cadparser: A learning approach of sequence modeling for b-rep cad

    Shengdi Zhou, Tianyi Tang, and Bin Zhou. Cadparser: A learning approach of sequence modeling for b-rep cad. In IJCAI, pages 1804–1812, 2023. 3

    2023

  44. [52]

    Minigpt-4: Enhancing vision-language understanding with advanced large language models

    Deyao Zhu, Jun Chen, Xiaoqian Shen, Xiang Li, and Mo- hamed Elhoseiny. Minigpt-4: Enhancing vision-language understanding with advanced large language models. arXiv preprint arXiv:2304.10592, 2023. 3

    2023 arXiv

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.