Pith. sign in

REVIEW 4 major objections 5 minor 74 references

GarmentX: Autoregressive Parametric Representations for High-Fidelity 3D Garment Generation

T0 review · 4 major / 5 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 Predicting garment parameters, not pattern edges, yields wearable 3D garments from a single image.

desk verdict A useful new dataset and a sensible parametric recipe, but the 'always valid' guarantee is under-specified and needs a fix before publication. read the letter →

arxiv 2504.20409 v1 pith:CUQTYYYY submitted 2025-04-29 cs.CV

Jingfeng Guo , Jinnan Chen , Weikai Chen , Zhenyu Sun , Lanjiong Li , Baozhu Zhao , Lingting Zhu , Xin Wang
show 1 more author
Qi Liu
This is my paper · ORCID
classification cs.CV
keywords 3Dgarmentgenerationsewingpatternpredictionmaskedautoregressivemodelparametricrepresentationimage-to-garmentdiffusiondatasetclothsimulation
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

GarmentX claims that image-to-garment methods fail because they predict low-level sewing-pattern geometry, an unconstrained space full of invalid configurations. The paper instead predicts a structured vector of high-level, editable garment parameters — neckline type, sleeve shape, length, flare — that is guaranteed to decode through GarmentCode into a valid, simulation-ready sewing pattern. On the CLOTH3D benchmark the method reports the best Chamfer distance (4.92) and point-to-surface distance (4.96), zero simulation failures, and 15-second inference, and it supports direct editing by changing parameters. A dataset of 378,682 parameter-image pairs, with photorealistic images synthesized from rendered garments, provides the training substrate. If these results hold, a single image becomes a practical entry point to production-grade editable digital clothing.

What carries the argument

The load-bearing object is the GarmentX parameter representation: a fixed-length vector of semantically meaningful garment attributes, normalized to $[-1,1]$ (continuous values linearly rescaled; $K$ discrete choices mapped to midpoints of $K$ equal intervals). Because this vector is a valid GarmentCode program, the difficult problem of predicting panel edges, stitches, and connectivity is removed from the network output. The predictor is a masked autoregressive model: DINOv2 extracts image tokens, an MAE-style encoder-decoder fuses them with partially masked parameter tokens, and a diffusion MLP denoises each token to its ground-truth parameter value; tokens are generated in a cosine-scheduled order and classifier-free guidance is applied. The data-construction pipeline — rendering multi-view images in Blender and photorealism-enhancing them with canny-conditioned ControlNet — supplies the paired supervision that makes the image-to-parameter map learnable.

What would settle it

Take a set of real garment photos with known ground-truth sewing patterns (or high-quality scans of the same garments), run GarmentX on them, and compare predicted parameters and simulated meshes against ground truth; if Chamfer or point-to-surface errors on real photos are much larger than on synthetic pairs, or if simulation failures reappear, the claimed validity guarantee and input-image alignment fail on real input. A cheaper check: uniformly sample random parameter vectors and verify that every one decodes to a valid pattern; any decoded self-intersection would refute the 'always valid' claim.

Watch

Extended reading notes

Core claim

The paper's central discovery is that constraining generation to a parametric garment space converts an ill-posed pattern-assembly problem into a tractable sequence-prediction problem. Each garment is represented by a normalized parameter vector, with continuous attributes linearly rescaled to $[-1,1]$ and discrete choices mapped to interval midpoints; because every vector in this space corresponds to a valid GarmentCode program, the decoded 2D sewing patterns always stitch up into open, single-layer, wearable 3D garments. A masked autoregressive transformer, conditioned on DINOv2 image tokens and using a diffusion denoiser for the continuous parameter values, predicts the vector; a cosine generation schedule and classifier-free guidance balance quality and condition adherence. Trained on the GarmentX dataset and evaluated on CLOTH3D, the model achieves Chamfer distance 4.92 and point-to-surface distance 4.96 with no simulation failures, outperforming pattern-based, deformation-based, and general 3D-generation baselines, and runs in 15 seconds.

Load-bearing premise

The load-bearing premise is that canny-conditioned ControlNet images made from rendered 3D garments preserve the geometric cues (folds, hemlines, collars, proportions) that the model needs, so the image-to-parameter mapping learned on synthetic pairs transfers to real photographs.

Editorial extensions

If this is right

  • Direct pattern predictors like SewFormer fail in 76.3% of simulations; GarmentX-style parameter prediction brings the failure rate to zero on the benchmark, making simulation robustness a property of the representation rather than of network luck.
  • Generated garments are open-structure and single-layer, so they can be draped onto arbitrary human bodies and animated, unlike closed or double-layered meshes produced by general image-to-3D models.
  • Because the output is a parameter vector, users can edit sleeves, collars, hoods, and even garment category after generation; editing is geometry-aware rather than a mesh deformation.
  • Autoregressive token generation with diffusion denoising reaches 15-second inference, about 60 times faster than deformation-based Garment3DGen, making interactive use plausible.
  • The model generalizes beyond its training distribution to sketches and uncommon garment types such as asymmetric one-shoulder dresses.

Reading between the lines

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

  • The dataset's reliance on canny-conditioned ControlNet as a realism bridge means the method's real-world ceiling is set by how faithfully that model preserves geometric cues; a testable extension is training on a small set of real photos with paired parameters to measure and close the gap.
  • A natural next step is text-conditioned generation: replacing the image-token condition with text embeddings would let the same parametric autoregressor produce garments from descriptions while retaining the validity guarantee.
  • The recipe of restricting the output space to a closed parametric manifold may transfer to other structured assets, such as furniture, buildings, or mechanical parts, where decoded outputs must satisfy hard physical constraints.
  • If the parameter space is truly closed under GarmentCode decoding, then random edits within normalized bounds should always stay valid; a simple stress test would verify this central promise.
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 / 5 minor

Summary. GarmentX proposes a single-image 3D garment generation framework that predicts a structured parametric representation compatible with GarmentCode, rather than directly predicting sewing patterns. A masked autoregressive model with a diffusion loss generates the parameters from DINOv2 image tokens; the parameters are then decoded into sewing patterns and simulated into 3D garments. The paper also introduces a 378,682-pair synthetic dataset built by rendering GarmentCodeData garments and applying canny-conditioned ControlNet, and reports state-of-the-art Chamfer distance (4.92) and point-to-surface distance (4.96) on CLOTH3D with zero simulation failures, along with ablations over the autoregressive backbone, CFG scale, and data construction choices.

Significance. The underlying idea is promising: constraining generation to a validated parametric space could eliminate the pattern failures of direct edge/stitch predictors while retaining editability. If the validity guarantee and the quantitative results hold, the 60x runtime improvement over deformation-based baselines and the large dataset would be valuable community contributions. The paper also provides substantial architectural detail and includes ablations over model choice, CFG scale, and data construction. However, the key guarantee is not yet substantiated because the post-processing of diffusion outputs is unspecified, and the evaluation lacks the statistical and protocol details needed to judge the state-of-the-art claim. The zero-failure and runtime claims are concrete and testable, and evaluation on an external benchmark (CLOTH3D) is a strength; the work is significant in potential but requires a careful revision.

major comments (4)
  1. [§3.3, Eq. (3)] The central guarantee that decoded sewing patterns 'always form valid, simulation-ready 3D garments' is not supported by the inference procedure as described. The reverse diffusion process produces continuous denoised samples, but the manuscript never specifies how these samples are constrained before the inverse transform N^{-1}: no clamping to [-1,1] for continuous parameters and no rounding/quantization to the nearest discrete midpoint introduced in Eq. (2) is reported. Without an explicit constraint, denoised outputs can lie outside the normalized range or between category midpoints, and GarmentCode would receive values outside the validated domain. Please state the exact post-processing (e.g., clamp plus nearest-midpoint quantization), evaluate its impact, or revise the 'always valid' claim.
  2. [§3.2] The conversion from GarmentCodeData's low-level attributes to the GarmentX representation is described only as a transformation; the paper does not provide the attribute inventory, the mapping rules from panel/stitch data to the parameter vector p, the determination of p_min/p_max for Eq. (1), or a validation procedure that filters parameter vectors that GarmentCode cannot decode. This underspecification is load-bearing because the dataset construction and the claimed guarantee that every sampled parameter vector yields a valid garment rest on it. Please provide the mapping details or an explicit reference to the GarmentCode API and any filtering steps.
  3. [§4.1, Table 1] The quantitative evaluation lacks essential context: the CLOTH3D test set size is not reported, no variance or error bars are given, and the protocol for producing input images from CLOTH3D (viewpoint, resolution, renderer, and whether images are original or newly rendered) is not documented. DressCode is mentioned as a baseline in §4.2 but does not appear in Table 1, and the 'Wearable' column is undefined. Please add these details and state how the input-image protocol affects the generalization-to-real-photos claim.
  4. [§3.2, Fig. 2] The data construction pipeline generates ControlNet-enhanced images from the same parametric representation that serves as prediction targets, so any geometric distortion introduced by the realism-enhancement step (e.g., altered folds, hemlines, collars) can bias the learned image-to-parameter mapping. The paper provides no check that ControlNet images are geometrically faithful to the original renders. A quantitative comparison of parameters predicted from ControlNet images versus the original renders would substantiate the claim that these synthetic pairs are faithful proxies for real photos; without this, the input-image alignment claim on real images is only weakly supported.
minor comments (5)
  1. [§3.2] Typos 'realistc' and 'compatability' should be corrected.
  2. [§3.3] The notation omega_s = s * lambda_cfg / S is ambiguous; clarify that s is the auto-regressive step index and S is the total number of steps.
  3. [Table 3] The caption contains the typo 'compartion' and should read 'comparison'.
  4. [Supplementary §6] The relationship between the 32 auto-regressive steps and the token count N is unclear; please clarify how the cosine schedule in Eq. (4) is applied when N differs from 32.
  5. [§4.2] DressCode is compared only qualitatively; either report its quantitative numbers in Table 1 or justify its exclusion from the quantitative comparison.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the derivation chain is self-contained and the key claims are tested against an external benchmark with an external decoder.

full rationale

The paper's central derivation is not circular. Garment parameters are normalized through Eq. 1 and Eq. 2, predicted by a masked autoregressive diffusion model through Eq. 3, and then projected back with N^{-1} before being decoded by GarmentCode. None of these steps defines the predicted output in terms of the input by construction; the mapping from image features to parameters is learned and evaluated on CLOTH3D, an external benchmark not used for training. GarmentCode and GarmentCodeData are external references used as a fixed decoder and data source, and the validity guarantee is inherited from decoding parameters that lie in the GarmentCode-validated domain, not from a self-referential fit. Self-citations appear only in related-work positioning (e.g., MAR-3D [11]) and are not load-bearing for the central claim. One non-circular support gap should be flagged: Section 3.3 asserts that 'strictly constrained parameter boundaries' eliminate simulation failures, but the paper does not specify how raw diffusion outputs are constrained (e.g., clamped to [-1,1] or quantized to discrete midpoints) before the inverse transform N^{-1}; this is a missing technical detail or an unsupported guarantee, not an equation-level circularity. Overall, the evaluation is external, the decoder is external, and there is no fitted-input-called-prediction or uniqueness-imported-from-authors step, so the circularity score is 0.

Assumptions & free parameters 1 free parameters · 4 assumptions · 1 invented entities

The central claims rest on the external GarmentCode decoder, the fidelity of the synthetic ControlNet images as training proxies, and the meaningfulness of the CLOTH3D benchmark. The only parameter directly tuned on the final evaluation set is the CFG scale. No new physical entities are introduced; the GarmentX representation is a designed intermediate representation rather than an empirical discovery.

free parameters (1)
  • CFG scale (lambda_cfg) = 3.0
    Selected by sweeping values 2.0 to 5.0 on the same CLOTH3D set used for the final comparison in Table 1, so the headline result is tuned on the benchmark.
assumptions (4)
  • domain assumption GarmentCode parameter ranges guarantee valid, simulation-ready sewing patterns whenever parameters fall within the allowed boundaries.
    Sections 3.1 and 3.3 invoke GarmentCode as a black-box decoder that turns any valid parameter vector into a valid sewing pattern; the paper provides no proof or failure analysis of this guarantee beyond citing the external work.
  • domain assumption Canny-conditioned ControlNet transformations preserve the garment geometry and semantic identity of the rendered image, so the synthesized photorealistic images remain valid renderings of the annotated parameter vector.
    Section 3.2 (Figure 2) relies on ControlNet to convert rendered images into photorealistic training pairs; if this changes shape cues, the image-parameter pairs are misaligned.
  • domain assumption CLOTH3D validation ground-truth meshes and rendered images are reliable references for measuring input-image alignment via Chamfer distance and point-to-surface distance.
    Section 4.1 uses CLOTH3D as the only quantitative benchmark without reporting test-set size, preprocessing, or how garment meshes are extracted.
  • domain assumption DINOv2 visual features carry sufficient garment shape and style information for the autoregressive model to predict parameters from a single image.
    Section 3.3 conditions generation on DINOv2 ViT-B/14 tokens; no analysis shows which features are used or that they are sufficient for fine-grained garment parameters.
invented entities (1)
  • GarmentX parameter representation
    purpose: Curated vector of garment attributes such as neckline, sleeve, length, and flare that serves as the intermediate prediction target and is decoded into sewing patterns through GarmentCode.
    The representation is constructed by converting GarmentCodeData attributes into a new schema; the paper does not provide independent evidence that this representation is complete or unbiased relative to original garment geometry, and its validity depends entirely on GarmentCode's decoder.

how reviews work

0 comments
Cite this review

Pith. "Pith review of GarmentX: Autoregressive Parametric Representations for High-Fidelity 3D Garment Generation." pith.science (2026). https://pith.science/paper/CUQTYYYY

@misc{pith2026250420409,
  author       = {Pith},
  title        = {Pith review of: GarmentX: Autoregressive Parametric Representations for High-Fidelity 3D Garment Generation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/CUQTYYYY}},
  note         = {Machine review of arXiv:2504.20409}
}
read the original abstract

This work presents GarmentX, a novel framework for generating diverse, high-fidelity, and wearable 3D garments from a single input image. Traditional garment reconstruction methods directly predict 2D pattern edges and their connectivity, an overly unconstrained approach that often leads to severe self-intersections and physically implausible garment structures. In contrast, GarmentX introduces a structured and editable parametric representation compatible with GarmentCode, ensuring that the decoded sewing patterns always form valid, simulation-ready 3D garments while allowing for intuitive modifications of garment shape and style. To achieve this, we employ a masked autoregressive model that sequentially predicts garment parameters, leveraging autoregressive modeling for structured generation while mitigating inconsistencies in direct pattern prediction. Additionally, we introduce GarmentX dataset, a large-scale dataset of 378,682 garment parameter-image pairs, constructed through an automatic data generation pipeline that synthesizes diverse and high-quality garment images conditioned on parametric garment representations. Through integrating our method with GarmentX dataset, we achieve state-of-the-art performance in geometric fidelity and input image alignment, significantly outperforming prior approaches. We will release GarmentX dataset upon publication.

Figures

Figures reproduced from arXiv: 2504.20409 by the authors.

Figure 1
Figure 1. GarmentX is an image-guided 3D garment generator that produces editable garment parameters, decodes valid sewing patterns, [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Automatic Data Construction Pipeline. We construct garment parameters-image pairs using ControlNet and Blender. 3. Method We propose GarmentX, a structured and parametric-driven framework for generating diverse, editable, and simulation￾ready 3D garments from a single input image. Central to our approach is a structured garment parameter represen￾tation that ensures physically valid and simulation-ready garment synt… view at source ↗
Figure 3
Figure 3. Overview of GarmentX. Taking a single image as input, GarmentX trains a masked autoregressive generation model directly upon our GarmentX representation, extracts condition image tokens via DINOv2, processes them through MAE encoder-decoder architec￾ture and diffusion MLP. The generated GarmentX representation is projected back to original scale and then reconstruct sewing patterns through GarmentCode, and finally s… view at source ↗
Figures from the paper (8 more)
Figure 4
Figure 4. Figure 4: Qualitative comparisons on CLOTH3D dataset. GarmentX produce wearable, open-structure, and complete 3D garments. Implementation Details. Input images are resized into 224×224 and processed by the DINOv2 ViT-B/14 model to extract condition tokens. Training completes in …
Figure 5
Figure 5. Figure 5: Qualitative comparisons with Trellis. GarmentX gen￾erates single-layer, open-structure 3D garments. completely closes when dealing with the T-shirt images. In contrast, our method produces single-layer, open-structure garments, which can be directly draped onto various…
Figure 6
Figure 6. Figure 6: Qualitative comparisons on CVDD dataset. GarmentX produce diverse, complex, and detailed 3D garments that match the input images well. top harnesses, and strapless dresses, which DressCode fails to produce. Additionally, as evidenced in [PITH_FULL_IMAGE:figures/full_f…
Figure 7
Figure 7. Figure 7: Qualitative comparisons with DressCode. GarmentX exhibits better generalization, enabling the creation of a variety of uncommon garments [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]
Figure 9
Figure 9. Figure 9: Qualitative ablation of different data construction methods. ControlNet was selected for its optimal balance between speed and quality [PITH_FULL_IMAGE:figures/full_fig_p008_9.png]
Figure 10
Figure 10. Figure 10: Segmentation result as Input. We use SAM2 to get the segmentation result and use it as the input to our model. 12 [PITH_FULL_IMAGE:figures/full_fig_p012_10.png]
Figure 11
Figure 11. Figure 11: Some failure cases. GarmentX fails when handling highly complex cases [PITH_FULL_IMAGE:figures/full_fig_p013_11.png]
Figure 12
Figure 12. Figure 12: More qualitative results. GarmentX can handle diverse input images with various styles and textures. It accurately captures geometric details in the input images and generates high fidelity, open-structure and simulation-ready 3D garments, facilitating downstream appl…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

74 extracted references · 50 canonical work pages

  1. [1]

    https://3dmd.com/

    3dMD. https://3dmd.com/. 2

    work page

  2. [2]

    https://www.artec3d.com/portable- 3d-scanners/

    Artec3D. https://www.artec3d.com/portable- 3d-scanners/. 2

    work page

  3. [3]

    https://clo3d.com/en/

    Clo3D. https://clo3d.com/en/. 2

    work page

  4. [4]

    CLOTH3D: Clothed 3D Humans

    Hugo Bertiche, Meysam Madadi, and Sergio Escalera. CLOTH3D: Clothed 3D Humans. In European Conference on Computer Vision, pages 344–359. Springer, 2020. 5

    work page 2020

  5. [5]

    Chat- Garment: Garment Estimation, Generation and Editing via Large Language Models

    Siyuan Bian, Chenghao Xu, Yuliang Xiu, Artur Grigorev, Zhen Liu, Cewu Lu, Michael J Black, and Yao Feng. Chat- Garment: Garment Estimation, Generation and Editing via Large Language Models. arXiv preprint arXiv:2412.17811,

    arXiv

  6. [6]

    Huiwen Chang, Han Zhang, Lu Jiang, Ce Liu, and William T. Freeman. MaskGIT: Masked Generative Image Transformer. In The IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2022. 3, 5

    work page 2022

  7. [7]

    Freeman, Michael Rubinstein, Yuanzhen Li, and Dilip Krishnan

    Huiwen Chang, Han Zhang, Jarred Barber, AJ Maschinot, Jose Lezama, Lu Jiang, Ming-Hsuan Yang, Kevin Mur- phy, William T. Freeman, Michael Rubinstein, Yuanzhen Li, and Dilip Krishnan. Muse: Text-To-Image Generation via Masked Generative Transformers, 2023. 3

    work page 2023

  8. [8]

    PanelFormer: Sewing Pattern Re- construction from 2D Garment Images

    Cheng-Hsiu Chen, Jheng-Wei Su, Min-Chun Hu, Chih-Yuan Yao, and Hung-Kuo Chu. PanelFormer: Sewing Pattern Re- construction from 2D Garment Images. InProceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, pages 454–463, 2024. 2

    work page 2024

Show all 74 references
  1. [9]

    Dihur: Diffusion- guided generalizable human reconstruction

    Jinnan Chen, Chen Li, and Gim Hee Lee. Dihur: Diffusion- guided generalizable human reconstruction. arXiv preprint arXiv:2411.11903, 2024. 1

    2024 arXiv

  2. [10]

    Gener- alizable human gaussians from single-view image

    Jinnan Chen, Chen Li, Jianfeng Zhang, Lingting Zhu, Buzhen Huang, Hanlin Chen, and Gim Hee Lee. Gener- alizable human gaussians from single-view image. arXiv preprint arXiv:2406.06050, 2024. 1

    2024 arXiv

  3. [11]

    Mar-3d: Pro- gressive masked auto-regressor for high-resolution 3d gener- ation

    Jinnan Chen, Lingting Zhu, Zeyu Hu, Shengju Qian, Yu- gang Chen, Xin Wang, and Gim Hee Lee. Mar-3d: Pro- gressive masked auto-regressor for high-resolution 3d gener- ation. arXiv preprint arXiv:2503.20519, 2025. 3

    2025 arXiv

  4. [12]

    SAR3D: Autoregressive 3D Object Gen- eration and Understanding via Multi-scale 3D VQV AE

    Yongwei Chen, Yushi Lan, Shangchen Zhou, Tengfei Wang, and Xingang Pan. SAR3D: Autoregressive 3D Object Gen- eration and Understanding via Multi-scale 3D VQV AE. In CVPR, 2025. 3

    2025

  5. [13]

    DrapeNet: Garment Generation and Self- Supervised Draping

    Luca De Luigi, Ren Li, Benoit Guillard, Mathieu Salzmann, and Pascal Fua. DrapeNet: Garment Generation and Self- Supervised Draping. In Proceedings of the IEEE/CVF Con- ference on Computer Vision and Pattern Recognition, pages 1451–1460, 2023. 2

    2023

  6. [14]

    Towards multi-pose guided virtual try-on network

    Haoye Dong, Xiaodan Liang, Xiaohui Shen, Bochao Wang, Hanjiang Lai, Jia Zhu, Zhiting Hu, and Jian Yin. Towards multi-pose guided virtual try-on network. In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), 2019. 1

    2019

  7. [15]

    Fw-gan: Flow-navigated warping gan for video virtual try-on

    Haoye Dong, Xiaodan Liang, Xiaohui Shen, Bowen Wu, Bing-Cheng Chen, and Jian Yin. Fw-gan: Flow-navigated warping gan for video virtual try-on. In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), 2019

    2019

  8. [16]

    Df-vton: Dense flow guided virtual try-on network

    Haoye Dong, Jun Liu, and Dong Huang. Df-vton: Dense flow guided virtual try-on network. In ICASSP 2024 - 2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pages 3175–3179, 2024. 1

    2024

  9. [17]

    Taming Transformers for High-Resolution Image Synthesis

    Patrick Esser, Robin Rombach, and Bjorn Ommer. Taming Transformers for High-Resolution Image Synthesis. In Pro- ceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 12873–12883, 2021. 3

    2021

  10. [18]

    Dart: Denoising autoregressive transformer for scalable text-to-image generation, 2025

    Jiatao Gu, Yuyang Wang, Yizhe Zhang, Qihang Zhang, Dinghuai Zhang, Navdeep Jaitly, Josh Susskind, and Shuangfei Zhai. Dart: Denoising autoregressive transformer for scalable text-to-image generation, 2025. 3

    2025

  11. [19]

    Vid2avatar-pro: Authentic avatar from videos in the wild via universal prior, 2025

    Chen Guo, Junxuan Li, Yash Kant, Yaser Sheikh, Shunsuke Saito, and Chen Cao. Vid2avatar-pro: Authentic avatar from videos in the wild via universal prior, 2025. 1

    2025

  12. [20]

    Masked Autoencoders Are Scal- able Vision Learners

    Kaiming He, Xinlei Chen, Saining Xie, Yanghao Li, Piotr Doll´ar, and Ross Girshick. Masked Autoencoders Are Scal- able Vision Learners. arXiv:2111.06377, 2021. 3

    2021 arXiv

  13. [21]

    Masked Autoencoders are Scal- able Vision Learners

    Kaiming He, Xinlei Chen, Saining Xie, Yanghao Li, Piotr Doll´ar, and Ross Girshick. Masked Autoencoders are Scal- able Vision Learners. In Proceedings of the IEEE/CVF con- ference on computer vision and pattern recognition , pages 16000–16009, 2022. 5

    2022

  14. [22]

    DressCode: Autoregressively Sewing and Generating Garments from Text Guidance

    Kai He, Kaixin Yao, Qixuan Zhang, Jingyi Yu, Lingjie Liu, and Lan Xu. DressCode: Autoregressively Sewing and Generating Garments from Text Guidance. arXiv preprint arXiv:2401.16465, 2024. 2, 6

    2024 arXiv

  15. [23]

    Classifier-Free Diffusion Guidance

    Jonathan Ho and Tim Salimans. Classifier-Free Diffusion Guidance. arXiv preprint arXiv:2207.12598, 2022. 5

    2022 arXiv

  16. [24]

    Gauhuman: Articulated gaus- sian splatting for real-time 3d human rendering

    Shoukang Hu and Ziwei Liu. Gauhuman: Articulated gaus- sian splatting for real-time 3d human rendering. arXiv preprint, 2023. 1

    2023

  17. [25]

    Humangif: Single-view human diffusion with generative prior

    Shoukang Hu, Takuya Narihira, Kazumi Fukuda, Ryosuke Sawata, Takashi Shibuya, and Yuki Mitsufuji. Humangif: Single-view human diffusion with generative prior. arXiv preprint arXiv:2502.12080, 2025. 1

    2025 arXiv

  18. [26]

    X- ray: A sequential 3d representation for generation

    Tao Hu, Wenhang Ge, Yuyang Zhao, and Gim Hee Lee. X- ray: A sequential 3d representation for generation. Advances in Neural Information Processing Systems , 37:136193– 136219, 2024. 1

    2024

  19. [27]

    FitDiT: Advancing the Authentic Garment Details for High-fidelity Virtual Try- on

    Boyuan Jiang, Xiaobin Hu, Donghao Luo, Qingdong He, Chengming Xu, Jinlong Peng, Jiangning Zhang, Chengjie Wang, Yunsheng Wu, and Yanwei Fu. FitDiT: Advancing the Authentic Garment Details for High-fidelity Virtual Try- on. arXiv preprint arXiv:2411.10499, 2024. 5

    2024 arXiv

  20. [28]

    NeuralTailor: Recon- structing Sewing Pattern Structures from 3D Point cClouds of Garments

    Maria Korosteleva and Sung-Hee Lee. NeuralTailor: Recon- structing Sewing Pattern Structures from 3D Point cClouds of Garments. ACM Transactions on Graphics (TOG), 41(4): 1–16, 2022. 2 9

    2022

  21. [29]

    Garment- Code: Programming Parametric Sewing Patterns

    Maria Korosteleva and Olga Sorkine-Hornung. Garment- Code: Programming Parametric Sewing Patterns. ACM Transactions on Graphics (TOG), 42(6):1–15, 2023. 3

    2023

  22. [30]

    GarmentCodeData: A Dataset of 3D Made-to-Measure Garments with Sewing Pat- terns

    Maria Korosteleva, Timur Levent Kesdogan, Fabian Kem- per, Stephan Wenninger, Jasmin Koller, Yuhan Zhang, Mario Botsch, and Olga Sorkine-Hornung. GarmentCodeData: A Dataset of 3D Made-to-Measure Garments with Sewing Pat- terns. arXiv preprint arXiv:2405.17609, 2024. 2

    2024 arXiv

  23. [31]

    Gar- mentDreamer: 3DGS Guided Garment Synthesis with Di- verse Geometry and Texture Details

    Boqian Li, Xuan Li, Ying Jiang, Tianyi Xie, Feng Gao, Huamin Wang, Yin Yang, and Chenfanfu Jiang. Gar- mentDreamer: 3DGS Guided Garment Synthesis with Di- verse Geometry and Texture Details. arXiv preprint arXiv:2405.12420, 2024. 2, 3

    2024 arXiv

  24. [32]

    Black, and Yuliang Xiu

    Boqian Li, Haiwen Feng, Zeyu Cai, Michael J. Black, and Yuliang Xiu. Etch: Generalizing body fitting to clothed hu- mans via equivariant tightness, 2025. 1

    2025

  25. [33]

    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 Process- ing Systems, 37:56424–56445, 2025. 3, 5

    2025

  26. [34]

    DiffAvatar: Simulation-ready garment optimization with differentiable simulation

    Yifei Li, Hsiao-yu Chen, Egor Larionov, Nikolaos Sarafi- anos, Wojciech Matusik, and Tuur Stuyck. DiffAvatar: Simulation-ready garment optimization with differentiable simulation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2024. 3

    2024

  27. [35]

    Triposg: High-fidelity 3d shape synthesis using large-scale rectified flow models

    Yangguang Li, Zi-Xin Zou, Zexiang Liu, Dehu Wang, Yuan Liang, Zhipeng Yu, Xingchao Liu, Yuan-Chen Guo, Ding Liang, Wanli Ouyang, et al. Triposg: High-fidelity 3d shape synthesis using large-scale rectified flow models. arXiv preprint arXiv:2502.06608, 2025. 1

    2025 arXiv

  28. [36]

    Towards Garment Sewing Pattern Recon- struction from a Single Image

    Lijuan Liu, Xiangyu Xu, Zhijie Lin, Jiabin Liang, and Shuicheng Yan. Towards Garment Sewing Pattern Recon- struction from a Single Image. ACM Transactions on Graph- ics (TOG), 42(6):1–15, 2023. 2, 6

    2023

  29. [37]

    ClotheDreamer: Text-Guided Garment Generation with 3D Gaussians

    Yufei Liu, Junshu Tang, Chu Zheng, Shijie Zhang, Jinkun Hao, Junwei Zhu, and Dongjin Huang. ClotheDreamer: Text-Guided Garment Generation with 3D Gaussians. arXiv preprint arXiv:2406.16815, 2024. 2, 3

    2024 arXiv

  30. [38]

    Dreamudf: Generating un- signed distance fields from a single image

    Yu-Tao Liu, Xuan Gao, Weikai Chen, Jie Yang, Xiaoxu Meng, Bo Yang, and Lin Gao. Dreamudf: Generating un- signed distance fields from a single image. ACM Transac- tions on Graphics (Proceedings of ACM SIGGRAPH 2024),

    2024

  31. [39]

    Geal: Generalizable 3d affordance learning with cross- modal consistency, 2024

    Dongyue Lu, Lingdong Kong, Tianxin Huang, and Gim Hee Lee. Geal: Generalizable 3d affordance learning with cross- modal consistency, 2024. 1

    2024

  32. [40]

    GarVerseLOD: High-Fidelity 3D Garment Reconstruction from a Single In-the-Wild Image using a Dataset with Levels of Details.ACM Transactions on Graph- ics (TOG), 43(6):1–12, 2024

    Zhongjin Luo, Haolin Liu, Chenghong Li, Wanghao Du, Zirong Jin, Wanhu Sun, Yinyu Nie, Weikai Chen, and Xi- aoguang Han. GarVerseLOD: High-Fidelity 3D Garment Reconstruction from a Single In-the-Wild Image using a Dataset with Levels of Details.ACM Transactions on Graph- ics (T...

    2024

  33. [41]

    Token-shuffle: Towards high-resolution image generation with autoregressive mod- els, 2025

    Xu Ma, Peize Sun, Haoyu Ma, Hao Tang, Chih-Yao Ma, Jialiang Wang, Kunpeng Li, Xiaoliang Dai, Yujun Shi, Xuan Ju, Yushi Hu, Artsiom Sanakoyeu, Felix Juefei-Xu, Ji Hou, Junjiao Tian, Tao Xu, Tingbo Hou, Yen-Cheng Liu, Zecheng He, Zijian He, Matt Feiszli, Peizhao Zhang, Pe- ter V...

    2025

  34. [42]

    T2I-Adapter: Learn- ing Adapters to Dig Out More Controllable Ability for Text- to-Image Diffusion Models

    Chong Mou, Xintao Wang, Liangbin Xie, Yanze Wu, Jian Zhang, Zhongang Qi, and Ying Shan. T2I-Adapter: Learn- ing Adapters to Dig Out More Controllable Ability for Text- to-Image Diffusion Models. InProceedings of the AAAI con- ference on artificial intelligence, pages 4296–4304...

    2024

  35. [43]

    AIpparel: A Large Multimodal Generative Model for Digital Garments

    Kiyohiro Nakayama, Jan Ackermann, Timur Levent Kes- dogan, Yang Zheng, Maria Korosteleva, Olga Sorkine- Hornung, Leonidas J Guibas, Guandao Yang, and Gordon Wetzstein. AIpparel: A Large Multimodal Generative Model for Digital Garments. arXiv preprint arXiv:2412.03937 ,

    arXiv

  36. [44]

    DINO2: Learning Robust Visual Features without Supervi- sion

    Maxime Oquab, Timoth ´ee Darcet, Th ´eo Moutakanni, Huy V o, Marc Szafraniec, Vasil Khalidov, Pierre Fernandez, Daniel Haziza, Francisco Massa, Alaaeldin El-Nouby, et al. DINO2: Learning Robust Visual Features without Supervi- sion. arXiv preprint arXiv:2304.07193, 2023. 5

    2023 arXiv

  37. [45]

    Scalable Diffusion Models with Transformers

    William Peebles and Saining Xie. Scalable Diffusion Models with Transformers. In Proceedings of the IEEE/CVF Inter- national Conference on Computer Vision, pages 4195–4205,

  38. [46]

    Com- putational Pattern Making from 3D Garment Models

    Nico Pietroni, Corentin Dumery, Raphael Falque, Mark Liu, Teresa A Vidal-Calleja, and Olga Sorkine-Hornung. Com- putational Pattern Making from 3D Garment Models. ACM Trans. Graph., 41(4):157–1, 2022. 2

    2022

  39. [47]

    SAM 2: Segment Anything in Images and Videos

    Nikhila Ravi, Valentin Gabeur, Yuan-Ting Hu, Ronghang Hu, Chaitanya Ryali, Tengyu Ma, Haitham Khedr, Roman R¨adle, Chloe Rolland, Laura Gustafson, et al. SAM 2: Segment Anything in Images and Videos. arXiv preprint arXiv:2408.00714, 2024. 9

    2024 arXiv

  40. [48]

    Garment3DGen: 3D Garment Stylization and Texture Generation

    Nikolaos Sarafianos, Tuur Stuyck, Xiaoyu Xiang, Yilei Li, Jovan Popovic, and Rakesh Ranjan. Garment3DGen: 3D Garment Stylization and Texture Generation. arXiv preprint arXiv:2403.18816, 2024. 2, 3, 6

    2024 arXiv

  41. [49]

    Denois- ing Diffusion Implicit Models, 2022

    Jiaming Song, Chenlin Meng, and Stefano Ermon. Denois- ing Diffusion Implicit Models, 2022. 5

    2022

  42. [50]

    WordRobe: Text-Guided Generation of Textured 3D Garments

    Astitva Srivastava, Pranav Manu, Amit Raj, Varun Jampani, and Avinash Sharma. WordRobe: Text-Guided Generation of Textured 3D Garments. InEuropean Conference on Com- puter Vision, pages 458–475. Springer, 2025. 2, 3

    2025

  43. [51]

    Autoregressive Model Beats Diffusion: Llama for Scalable Image Generation

    Peize Sun, Yi Jiang, Shoufa Chen, Shilong Zhang, Bingyue Peng, Ping Luo, and Zehuan Yuan. Autoregressive Model Beats Diffusion: Llama for Scalable Image Generation. arXiv preprint arXiv:2406.06525, 2024. 3

    2024 arXiv

  44. [52]

    LGM: Large Multi-View Gaus- sian Model for High-Resolution 3D Content Creation

    Jiaxiang Tang, Zhaoxi Chen, Xiaokang Chen, Tengfei Wang, Gang Zeng, and Ziwei Liu. LGM: Large Multi-View Gaus- sian Model for High-Resolution 3D Content Creation. In European Conference on Computer Vision , pages 1–18. Springer, 2025. 2, 6

    2025

  45. [53]

    Visual Autoregressive Modeling: Scalable Image Generation via Next-Scale Prediction, 2024

    Keyu Tian, Yi Jiang, Zehuan Yuan, Bingyue Peng, and Li- wei Wang. Visual Autoregressive Modeling: Scalable Image Generation via Next-Scale Prediction, 2024. 3

    2024

  46. [54]

    TripoSR: Fast 3D Object Reconstruction from a Single Image

    Dmitry Tochilkin, David Pankratz, Zexiang Liu, Zixuan Huang, Adam Letts, Yangguang Li, Ding Liang, Christian Laforte, Varun Jampani, and Yan-Pei Cao. TripoSR: Fast 3D Object Reconstruction from a Single Image. arXiv preprint arXiv:2403.02151, 2024. 2 10

    2024 arXiv

  47. [55]

    Gomez, Lukasz Kaiser, and Illia Polosukhin

    Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszko- reit, Llion Jones, Aidan N. Gomez, Lukasz Kaiser, and Illia Polosukhin. Attention Is All You Need, 2023. 3

    2023

  48. [56]

    Disentangled clothed avatar generation from text de- scriptions, 2023

    Jionghao Wang, Yuan Liu, Zhiyang Dou, Zhengming Yu, Yongqing Liang, Xin Li, Wenping Wang, Rong Xie, and Li Song. Disentangled clothed avatar generation from text de- scriptions, 2023. 1

    2023

  49. [57]

    Fresa: Feedforward reconstruc- tion of personalized skinned avatars from few images, 2025

    Rong Wang, Fabian Prada, Ziyan Wang, Zhongshi Jiang, Chengxiang Yin, Junxuan Li, Shunsuke Saito, Igor San- testeban, Javier Romero, Rohan Joshi, Hongdong Li, Jason Saragih, and Yaser Sheikh. Fresa: Feedforward reconstruc- tion of personalized skinned avatars from few images, 2025. 1

    2025

  50. [58]

    Structured 3D Latents for Scalable and Versatile 3D Generation

    Jianfeng Xiang, Zelong Lv, Sicheng Xu, Yu Deng, Ruicheng Wang, Bowen Zhang, Dong Chen, Xin Tong, and Jiaolong Yang. Structured 3D Latents for Scalable and Versatile 3D Generation. arXiv preprint arXiv:2412.01506, 2024. 2, 6

    2024 arXiv

  51. [59]

    Dreamvton: Customizing 3d virtual try-on with personalized diffusion models, 2024

    Zhenyu Xie, Haoye Dong, Yufei Gao, Zehua Ma, and Xiao- dan Liang. Dreamvton: Customizing 3d virtual try-on with personalized diffusion models, 2024. 1

    2024

  52. [60]

    InstantMesh: Effi- cient 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: Effi- cient 3D Mesh Generation from a Single Image with Sparse-View Large Reconstruction Models. arXiv preprint arXiv:2404.07191, 2024. 2, 6

    2024 arXiv

  53. [61]

    High quality human image animation using regional supervision and mo- tion blur condition, 2024

    Zhongcong Xu, Chaoyue Song, Guoxian Song, Jianfeng Zhang, Jun Hao Liew, Hongyi Xu, You Xie, Linjie Luo, Gu- osheng Lin, Jiashi Feng, and Mike Zheng Shou. High quality human image animation using regional supervision and mo- tion blur condition, 2024. 1

    2024

  54. [62]

    Magicinfinite: Generating infinite talking videos with your words and voice, 2025

    Hongwei Yi, Tian Ye, Shitong Shao, Xuancheng Yang, Jiantong Zhao, Hanzhong Guo, Terrance Wang, Qingyu Yin, Zeke Xie, Lei Zhu, Wei Li, Michael Lingelbach, and Daquan Zhou. Magicinfinite: Generating infinite talking videos with your words and voice, 2025. 1

    2025

  55. [63]

    Surf-d: High-quality surface generation for arbitrary topologies using diffusion models

    Zhengming Yu, Zhiyang Dou, Xiaoxiao Long, Cheng Lin, Zekun Li, Yuan Liu, Norman M ¨uller, Taku Komura, Marc Habermann, Christian Theobalt, et al. Surf-d: High-quality surface generation for arbitrary topologies using diffusion models. arXiv preprint arXiv:2311.17050, 2023. 2

    2023 arXiv

  56. [64]

    Paint3D: Paint Anything 3D with Lighting-Less Texture Diffusion Models

    Xianfang Zeng, Xin Chen, Zhongqi Qi, Wen Liu, Zibo Zhao, Zhibin Wang, Bin Fu, Yong Liu, and Gang Yu. Paint3D: Paint Anything 3D with Lighting-Less Texture Diffusion Models. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 4252–4262,

  57. [65]

    Lagem: A large geometry model for 3d representation learning and diffusion, 2024

    Biao Zhang and Peter Wonka. Lagem: A large geometry model for 3d representation learning and diffusion, 2024. 1

    2024

  58. [66]

    Adding conditional control to text-to-image diffusion models

    Lvmin Zhang, Anyi Rao, and Maneesh Agrawala. Adding conditional control to text-to-image diffusion models. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 3836–3847, 2023. 2, 4

    2023

  59. [67]

    DreamMat: High-quality PBR Ma- terial Generation with Geometry-and Light-aware Diffusion Models

    Yuqing Zhang, Yuan Liu, Zhiyu Xie, Lei Yang, Zhongyuan Liu, Mengzhou Yang, Runze Zhang, Qilong Kou, Cheng Lin, Wenping Wang, et al. DreamMat: High-quality PBR Ma- terial Generation with Geometry-and Light-aware Diffusion Models. ACM Transactions on Graphics (TOG), 43(4):1–18,

  60. [68]

    DI-PCG: Diffusion-based Efficient Inverse Pro- cedural Content Generation for High-quality 3D Asset Cre- ation

    Wang Zhao, Yan-Pei Cao, Jiale Xu, Yuejiang Dong, and Ying Shan. DI-PCG: Diffusion-based Efficient Inverse Pro- cedural Content Generation for High-quality 3D Asset Cre- ation. arXiv preprint arXiv:2412.15200, 2024. 1

    2024 arXiv

  61. [69]

    Muma: 3d pbr texturing via multi-channel multi-view generation and agentic post- processing

    Lingting Zhu, Jingrui Ye, Runze Zhang, Zeyu Hu, Yingda Yin, Lanjiong Li, Jinnan Chen, Shengju Qian, Xin Wang, Qingmin Liao, and Lequan Yu. Muma: 3d pbr texturing via multi-channel multi-view generation and agentic post- processing. arXiv preprint arXiv:2503.18461, 2025. 1

    2025 arXiv

  62. [70]

    Idol: Instant photorealistic 3d human creation from a single image, 2024

    Yiyu Zhuang, Jiaxi Lv, Hao Wen, Qing Shuai, Ailing Zeng, Hao Zhu, Shifeng Chen, Yujiu Yang, Xun Cao, and Wei Liu. Idol: Instant photorealistic 3d human creation from a single image, 2024. 1 11 GarmentX: Autoregressive Parametric Representations for High-Fidelity 3D Garment Gen...

    2024

  63. [71]

    This design choice facilitates effective modeling of long-range dependencies in 3D geom- etry

    More Implementation Details The MAR component adopts a Transformer-based encoder- decoder architecture, featuring 16 layers in both the encoder and decoder paths, 16 attention heads per layer, and an em- bedding dimension of 768. This design choice facilitates effective modeli...

  64. [72]

    More Qualitative Results Here we show more qualitative results of GarmentX. As shown in the first three rows of Figure 12, our framework successfully generates various types of lower garments, ac- curately distinguishing between full-length pants (a, b, c, d), shorts (e), mini...

  65. [73]

    For clothed human inputs as shown in Figure 10, we employ SAM2 to segment for preprocessing

    Future Works The current GarmentX implementation processes only human-free garment images due to our automatic data gen- eration pipeline’s human-free design. For clothed human inputs as shown in Figure 10, we employ SAM2 to segment for preprocessing. Planned improvements incl...

  66. [74]

    Failure Cases While GarmentX supports a wide range of modern gar- ment categories, Figure 11 reveals failure cases in process- ing highly complex designs—particularly those contain- ing complex decorations, like oversized chest-attached bow knots and petal-like designs. For th...

Pith tools

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