Pith. sign in

REVIEW 5 major objections 6 minor 55 references

JADE: Joint-aware Latent Diffusion for 3D Human Generative Modeling

T0 review · 5 major / 6 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read Splitting a 3D body into 24 joint tokens, each with a skeleton position and local surface features, yields both accurate reconstruction and fine-grained editing.

desk verdict JADE's joint-aware latent decomposition is a real novelty, but the paper's central fine-grained control claim rests on an unverified locality assumption that needs to be enforced or tested quantitatively. read the letter →

arxiv 2412.20470 v1 pith:BIYIJWA5 submitted 2024-12-29 cs.CV

Haorui Ji , Rong Wang , Taojun Lin , Hongdong Li This is my paper · ORCID
classification cs.CV
keywords 3Dhumangenerationjoint-awarelatentrepresentationdiffusionpointcloudautoencoderdisentangledbodyeditingtransformer
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

This paper tries to show that a 3D human body latent representation can be simultaneously expressive and semantically controllable if the body is dispatched into joint-wise tokens rather than encoded as one holistic code. Each of the 24 tokens carries two parts: an extrinsic joint position that encodes skeleton structure and posture, and an intrinsic feature that encodes local surface geometry around that joint. A transformer autoencoder learns this factorization, and a cascaded pair of diffusion models first samples the skeleton structure and then generates per-joint surface features conditioned on it. The claim matters because the same representation supports accurate reconstruction, joint-level editing, and diverse unconditional generation, which earlier holistic or part-segmentation-based models do not achieve together.

What carries the argument

The central object is the joint-aware latent token sequence. A PointNet encoder produces one global feature, an MLP splits it into 24 tokens, a learnable positional embedding anchors each token to a body region, and transformer attention mixes information across joints before each token is projected into an extrinsic joint position e_i and an intrinsic geometry feature h_i. The decoder concatenates extrinsics and intrinsics with the shared positional embedding and reconstructs the point cloud. Three losses carry the learning: vertex and joint reconstruction, a cross-swap disentanglement loss taken from prior work, and a KL prior that regularizes the intrinsic space. The cascaded diffusion pipeline then realizes the factorization p(X) = product over joints of p(e_i) p(h_i | e_i) by training one diffusion on extrinsics and a second, DiT-based diffusion on intrinsics conditioned on the generated skeleton.

What would settle it

Swap the intrinsic features of a single token between two different poses of the same subject and reconstruct: if the token has drifted from its anatomical joint, the swapped body will show geometry artifacts leaking across unrelated body regions. A complementary check is to perturb each intrinsic feature independently and map the surface vertices that move, then test whether those vertex sets stay fixed around one joint across poses.

Watch

Extended reading notes

Core claim

On its own terms, JADE claims that a joint-aware latent representation, learned from surface point clouds, reaches the best reconstruction accuracy among compared methods, with mean per-vertex error 5.47 on DFAUST and 12.85 on SPRING, while keeping the latent space structured enough for direct manipulation. The extrinsic/intrinsic split is enforced by a cross-consistency loss that swaps intrinsic features between two poses of the same subject and requires the decoder to reconstruct each pose, plus joint-position supervision on the extrinsic outputs. Generation is treated as two cascaded conditional diffusions: one transformer-based model learns p(E) over joint positions, and a DiT-based model learns p(H|E) with the sampled skeleton as an adaptive conditioning signal. The reported generation metrics (APD 14.95 and self-intersection rate 1.05 on AMASS) are positioned as a balance between diversity and realism.

Load-bearing premise

The 24 learned tokens are assumed to correspond consistently to the same anatomical joints across all identities and poses, but that correspondence is only encouraged by joint supervision and positional embeddings, never explicitly enforced.

Editorial extensions

If this is right

  • Joint-level controllability follows directly: editing or interpolating a token's extrinsics changes posture while keeping local shape, and changing intrinsics alters local geometry without moving the skeleton.
  • The same latent can support identity transfer between subjects by exchanging intrinsic features while keeping extrinsics, as shown in the shape-transfer experiments.
  • Because the representation needs no part segmentation labels, it can be trained on registered human scans and ported to new fixed-topology human datasets.
  • Cascaded sampling from p(E) then p(H|E) yields unconditional generation with both skeletal plausibility and local detail, with reported self-intersection rates lower than several diffusion baselines.

Reading between the lines

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

  • A direct extension the authors do not test: quantify token-to-joint stability by perturbing each intrinsic feature and measuring whether the affected surface vertices cluster around the corresponding anatomical joint across many identities and poses.
  • If the extrinsic–intrinsic split is as clean as claimed, the second-stage diffusion could be re-purposed for pose-conditioned shape generation by conditioning on an external skeleton rather than a sampled one, which would test the factorization's generality.
  • The representation's point-cloud form suggests it could combine with image-based conditioning or differentiable rendering for textured generation, though the paper only sketches that as future work.
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

5 major / 6 minor

Summary. The paper presents JADE, a generative framework for 3D human body point clouds. The core idea is a joint-aware latent representation that factorizes a body into 24 joint tokens, each split into an extrinsic component (joint position) and an intrinsic feature vector intended to encode local surface geometry. A transformer-based autoencoder learns this representation using reconstruction, cross-disentanglement, and KL-prior losses, and a cascaded diffusion pipeline first models p(E), then p(H|E). The paper reports reconstruction MPVPE on DFAUST and SPRING, generation APD and SI on AMASS, and qualitative editing examples.

Significance. If the locality and disentanglement of the intrinsic tokens were established, JADE would be a useful step toward interpretable and controllable generative human modeling, with reconstruction accuracy that is competitive with or better than several learned baselines. The paper also provides comparisons to multiple methods and an ablation study. However, the central controllability claim is supported only by a single qualitative example, and several quantitative claims would be strengthened by error bars and a direct evaluation of editing fidelity. The cascaded diffusion contribution is not ablated, and key hyperparameters are omitted.

major comments (5)
  1. [§3.3, Fig. 1] The claim that each intrinsic token h_i encodes local surface geometry near joint i is not enforced by the architecture or the losses. The Tokenization Network maps the entire point cloud through a global PointNet feature and splits it via an MLP, and Lrec (Eq. 6), Ldis (Eq. 7), and Lprior (Eq. 8) are all set-level or global objectives. Nothing prevents the network from learning redundant global shape codes in H, and the only evidence for locality is the qualitative interpolation in Fig. 5. To support the fine-grained controllability claim, the paper should either enforce locality in the encoder (e.g., by aggregating features from local surface neighborhoods around each joint) or provide a quantitative per-joint editing test that shows a single-token operation affects only the corresponding body region.
  2. [§4.3, §4.5] The paper's central contribution is joint-level controllability, but the editing evaluation is entirely qualitative. Section 4.3 shows one interpolation example, and the ablation text claims that omitting Ljoint or Ldis makes "the quality of editable human shapes degrade significantly," yet Table 3 reports only MPVPE and no editing metric. A quantitative editing protocol is needed, such as measuring per-joint or per-part displacement after swapping or interpolating a single intrinsic token, and reporting whether changes are localized to that joint's surface region. Without such evidence, the main novelty over holistic latent methods remains unverified.
  3. [§4.1.3, Eqs. (5)–(8)] The loss weights lambda_j, lambda_c, and lambda_kl are never reported. Since Table 3 varies the presence of Ljoint and Ldis, and Lprior directly shapes the latent space that the diffusion models, these values are necessary for reproducibility and for interpreting the ablation results. Please report the exact values used in all experiments.
  4. [§4.2, §4.4, Tables 1–2] All quantitative results appear to be single runs with no error bars or statistical significance tests. The DFAUST reconstruction improvement over SemanticHuman is 0.23 mm (5.47 vs 5.70), and the generation metrics place JADE in the middle of the compared methods; without multiple seeds and variance estimates, it is difficult to determine whether the reported differences are meaningful. At minimum, report mean and standard deviation over several training runs, or justify why a single run is sufficient.
  5. [§3.4, §4.4] The cascaded diffusion pipeline is a stated contribution, but it is not ablated. No experiment compares against a single diffusion model over the joint latent Z (or over the concatenation of E and H) or against sampling from the learned priors directly. Since Section 3.4 motivates the cascade by the prior-hole problem and the factorization in Eq. (3), the paper should demonstrate that the cascade improves generation quality relative to a non-cascaded alternative.
minor comments (6)
  1. [Abstract] The abstract contains a typo: "fined-grained" should be "fine-grained".
  2. [§2.2] The text mentions "LION [50], SLIDE [50]" but both are cited as reference [50], and [50] is the LION paper; the SLIDE method is not defined in the bibliography. Please correct the citation or add the missing reference.
  3. [Eq. (3)] The notation p(X) = product_i p(z_i) is loose, because X is a point cloud and the factorization is over latent tokens, not directly over surface points. Rephrase to clarify that this is an assumption on the latent representation rather than an exact factorization of the point cloud distribution.
  4. [Table 3] The two rows in "Variant 3" are ambiguous: the reader cannot tell which loss is removed in each row. Label the rows explicitly, for example "w/o Ljoint" and "w/o Ldis".
  5. [§4.1.3, Eq. (6)] Please state how the ground-truth joint locations J in the reconstruction loss are obtained for each dataset (e.g., from SMPL parameters of the registered meshes), since the autoencoder operates on surface point clouds rather than on a parametric model.
  6. [Fig. 4] The color-coded error maps in Figure 4 would benefit from an explicit color scale or a description of the error range, so that the qualitative comparison is interpretable.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: reconstruction and generation results are benchmarked externally; the joint-aware latent is learned, not defined in terms of its claimed outputs.

full rationale

JADE's derivation is not circular. The latent representation is trained to reconstruct held-out meshes, with the autoencoder's reconstruction accuracy evaluated on test splits of DFAUST and SPRING (Table 1) and its generation quality evaluated on AMASS against external baselines (Table 2). The cascaded diffusion pipeline explicitly implements the factorization of Eq. 3, but the paper does not present the successful generation as a derivation from that factorization; it presents it as an empirical result on a public dataset. The interpolation and editing results are qualitative checks that the model obeys its own disentanglement objective Ldis (Eq. 7), which is a sanity check rather than an independent validation, but it is not a fitted parameter renamed as a prediction. The two self-citations ([15], [17]) are for background and an architecture detail and are not load-bearing. The unverified assumption that each intrinsic token h_i encodes surface geometry local to its joint is a correctness and soundness risk, not a circularity, because no equation reduces the claimed controllability to its own input by construction.

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

The central method relies on a set of hand-chosen hyperparameters and domain assumptions. The three loss weights in Equations 6 through 8 are never numerically specified, which is a significant reproducibility gap. The main domain assumptions are the fixed-topology point cloud representation, the per-joint independence factorization, and the implicit token-to-joint correspondence. No new physical entities are introduced.

free parameters (4)
  • lambda_j (joint loss weight) = not reported
    Weight for the joint position reconstruction term in Equation 6. Its value is never given, yet it is central to making the extrinsics match anatomical joints.
  • lambda_c (cross disentanglement loss weight) = not reported
    Weight for the swap reconstruction loss in Equation 7, which creates the extrinsic/intrinsic disentanglement claimed by the paper. No value or schedule is provided.
  • lambda_kl (prior loss weight) = not reported
    Weight for the KL divergence to a standard normal in Equation 8, which shapes the intrinsic latent space for generation. The value is not stated.
  • intrinsic feature dimension D_h = 128
    Chosen by ablation, with 128 giving the best MPVPE on validation. This is a hand-picked hyperparameter, not derived from theory.
assumptions (4)
  • domain assumption Human body surfaces can be represented as fixed-topology point clouds of N points with consistent connectivity across all subjects.
    Section 3.1 states the input is a set of surface point clouds with consistent connectivity, which is true for registered datasets like DFAUST, SPRING, and AMASS but limits application to non-registered or arbitrary-topology scans.
  • domain assumption The shape distribution factorizes as p(X) = product over joints of p(e_i) * p(h_i | e_i), with conditional independence across joints given extrinsics.
    Equation 3 in Section 3.2 is the foundation of the cascaded diffusion pipeline. If joint surface features are strongly coupled across distant joints, this factorization will not capture the full distribution.
  • domain assumption Swapping intrinsic features between two poses of the same subject and reconstructing with f_dec(E_i, H_j) provides a valid disentanglement signal.
    Section 3.3.1 defines the cross consistency loss L_cross. It assumes that intrinsics carry identity and local geometry while extrinsics carry pose, an assumption that may break under pose-dependent soft tissue deformations.
  • ad hoc to paper Each learned joint token remains semantically anchored to a fixed anatomical joint across all shapes.
    The tokenization network splits a global PointNet feature into tokens and adds a shared learnable positional embedding; no structural constraint enforces token-to-joint correspondence other than the Ljoint supervision on the output extrinsics. This is an assumption the paper does not verify directly.

how reviews work

0 comments
Cite this review

Pith. "Pith review of JADE: Joint-aware Latent Diffusion for 3D Human Generative Modeling." pith.science (2026). https://pith.science/paper/BIYIJWA5

@misc{pith2026241220470,
  author       = {Pith},
  title        = {Pith review of: JADE: Joint-aware Latent Diffusion for 3D Human Generative Modeling},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/BIYIJWA5}},
  note         = {Machine review of arXiv:2412.20470}
}
read the original abstract

Generative modeling of 3D human bodies have been studied extensively in computer vision. The core is to design a compact latent representation that is both expressive and semantically interpretable, yet existing approaches struggle to achieve both requirements. In this work, we introduce JADE, a generative framework that learns the variations of human shapes with fined-grained control. Our key insight is a joint-aware latent representation that decomposes human bodies into skeleton structures, modeled by joint positions, and local surface geometries, characterized by features attached to each joint. This disentangled latent space design enables geometric and semantic interpretation, facilitating users with flexible controllability. To generate coherent and plausible human shapes under our proposed decomposition, we also present a cascaded pipeline where two diffusions are employed to model the distribution of skeleton structures and local surface geometries respectively. Extensive experiments are conducted on public datasets, where we demonstrate the effectiveness of JADE framework in multiple tasks in terms of autoencoding reconstruction accuracy, editing controllability and generation quality compared with existing methods.

Figures

Figures reproduced from arXiv: 2412.20470 by the authors.

Figure 1
Figure 1. Overview of our joint-aware latent representation. We [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. A visual illustration of the autoencoder architecture that is used to train our joint-aware latent representation as well as its training [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. A visual illustration of the diffusion pipeline, where two cascaded diffusions are presented, one for extrinsic parameters [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (2 more)
Figure 5
Figure 5. Figure 5: Qualitative example of transferring a given human char [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 4
Figure 4. Figure 4: Qualitative visualization results on DFAUST dataset, [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

55 extracted references · 33 canonical work pages

  1. [1]

    A stochastic con- ditioning scheme for diverse human motion prediction

    Sadegh Aliakbarian, Fatemeh Sadat Saleh, Mathieu Salz- mann, Lars Petersson, and Stephen Gould. A stochastic con- ditioning scheme for diverse human motion prediction. In Proceedings of the IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition, pages 5223–5232, 2020. 6

    work page 2020

  2. [2]

    Scape: shape completion and animation of people

    Dragomir Anguelov, Praveen Srinivasan, Daphne Koller, Se- bastian Thrun, Jim Rodgers, and James Davis. Scape: shape completion and animation of people. In ACM SIGGRAPH 2005 Papers, pages 408–416. 2005. 1, 2

    work page 2005

  3. [3]

    Keep it smpl: Automatic estimation of 3d human pose and shape from a single image

    Federica Bogo, Angjoo Kanazawa, Christoph Lassner, Peter Gehler, Javier Romero, and Michael J Black. Keep it smpl: Automatic estimation of 3d human pose and shape from a single image. In Computer Vision–ECCV 2016: 14th Euro- pean Conference, Amsterdam, The Netherlands, October 11- 14, 2016, Proceedings, Part V 14, pages 561–578. Springer,

    2016

  4. [4]

    Dynamic faust: Registering human bod- ies in motion

    Federica Bogo, Javier Romero, Gerard Pons-Moll, and Michael J Black. Dynamic faust: Registering human bod- ies in motion. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 6233–6242,

  5. [5]

    Neural 3d mor- phable models: Spiral convolutional networks for 3d shape representation learning and generation

    Giorgos Bouritsas, Sergiy Bokhnyak, Stylianos Ploumpis, Michael Bronstein, and Stefanos Zafeiriou. Neural 3d mor- phable models: Spiral convolutional networks for 3d shape representation learning and generation. In Proceedings of the IEEE/CVF international conference on computer vision, pages 7213–7222, 2019. 2, 7

    work page 2019

  6. [6]

    Multi-person 3d pose estimation and tracking in sports

    Lewis Bridgeman, Marco V olino, Jean-Yves Guillemaut, and Adrian Hilton. Multi-person 3d pose estimation and tracking in sports. In Proceedings of the IEEE/CVF confer- ence on computer vision and pattern recognition workshops, pages 0–0, 2019. 1

    2019

  7. [7]

    Diffusion models beat gans on image synthesis

    Prafulla Dhariwal and Alexander Nichol. Diffusion models beat gans on image synthesis. Advances in neural informa- tion processing systems, 34:8780–8794, 2021. 2

    2021

  8. [8]

    An image is worth 16x16 words: Trans- formers for image recognition at scale

    Alexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn, Xiaohua Zhai, Thomas Unterthiner, Mostafa Dehghani, Matthias Minderer, Georg Heigold, Syl- vain Gelly, et al. An image is worth 16x16 words: Trans- formers for image recognition at scale. arXiv preprint arXiv:2010.11929, 2020. 1

    arXiv 2010

Show all 55 references
  1. [9]

    Diffuseq: Sequence to sequence text generation with diffusion models

    Shansan Gong, Mukai Li, Jiangtao Feng, Zhiyong Wu, and LingPeng Kong. Diffuseq: Sequence to sequence text generation with diffusion models. arXiv preprint arXiv:2210.08933, 2022. 2

    2022 arXiv

  2. [10]

    Diffusionbert: Improving gener- ative masked language models with diffusion models

    Zhengfu He, Tianxiang Sun, Kuanning Wang, Xuanjing Huang, and Xipeng Qiu. Diffusionbert: Improving gener- ative masked language models with diffusion models. arXiv preprint arXiv:2211.15029, 2022. 2

    2022 arXiv

  3. [11]

    Spaghetti: Editing implicit shapes through part aware generation.ACM Transactions on Graph- ics (TOG), 41(4):1–20, 2022

    Amir Hertz, Or Perel, Raja Giryes, Olga Sorkine-Hornung, and Daniel Cohen-Or. Spaghetti: Editing implicit shapes through part aware generation.ACM Transactions on Graph- ics (TOG), 41(4):1–20, 2022. 3

    2022

  4. [12]

    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

    2020

  5. [13]

    Video dif- fusion models

    Jonathan Ho, Tim Salimans, Alexey Gritsenko, William Chan, Mohammad Norouzi, and David J Fleet. Video dif- fusion models. arXiv:2204.03458, 2022. 2

    2022 arXiv

  6. [14]

    Lrm: Large reconstruction model for single image to 3d

    Yicong Hong, Kai Zhang, Jiuxiang Gu, Sai Bi, Yang Zhou, Difan Liu, Feng Liu, Kalyan Sunkavalli, Trung Bui, and Hao Tan. Lrm: Large reconstruction model for single image to 3d. arXiv preprint arXiv:2311.04400, 2023. 3

    2023 arXiv

  7. [15]

    Unsu- pervised 3d pose estimation with non-rigid structure-from- motion modeling

    Haorui Ji, Hui Deng, Yuchao Dai, and Hongdong Li. Unsu- pervised 3d pose estimation with non-rigid structure-from- motion modeling. In Proceedings of the IEEE/CVF Win- ter Conference on Applications of Computer Vision , pages 3314–3323, 2024. 5

    2024

  8. [16]

    Disentangled human body embedding based on deep hierar- chical neural network

    Boyi Jiang, Juyong Zhang, Jianfei Cai, and Jianmin Zheng. Disentangled human body embedding based on deep hierar- chical neural network. IEEE transactions on visualization and computer graphics, 26(8):2560–2575, 2020. 1, 2

    2020

  9. [17]

    Golfpose: Golf swing analyses with a monocular camera based human pose estimation

    Zhongyu Jiang, Haorui Ji, Samuel Menaker, and Jenq-Neng Hwang. Golfpose: Golf swing analyses with a monocular camera based human pose estimation. In 2022 IEEE In- ternational Conference on Multimedia and Expo Workshops (ICMEW), pages 1–6. IEEE, 2022. 1

    2022

  10. [18]

    Salad: Part-level latent diffusion for 3d shape gen- eration and manipulation

    Juil Koo, Seungwoo Yoo, Minh Hieu Nguyen, and Minhyuk Sung. Salad: Part-level latent diffusion for 3d shape gen- eration and manipulation. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 14441– 14451, 2023. 3

    2023

  11. [19]

    Editvae: Unsupervised parts-aware controllable 3d point cloud shape generation

    Shidi Li, Miaomiao Liu, and Christian Walder. Editvae: Unsupervised parts-aware controllable 3d point cloud shape generation. In Proceedings of the AAAI Conference on Arti- ficial Intelligence, pages 1386–1394, 2022. 3

    2022

  12. [20]

    Diffusion-lm improves control- lable text generation

    Xiang Li, John Thickstun, Ishaan Gulrajani, Percy S Liang, and Tatsunori B Hashimoto. Diffusion-lm improves control- lable text generation. Advances in Neural Information Pro- cessing Systems, 35:4328–4343, 2022. 2

    2022

  13. [21]

    Zero-1-to-3: Zero-shot one image to 3d object

    Ruoshi Liu, Rundi Wu, Basile Van Hoorick, Pavel Tok- makov, Sergey Zakharov, and Carl V ondrick. Zero-1-to-3: Zero-shot one image to 3d object. In Proceedings of the IEEE/CVF International Conference on Computer Vision , pages 9298–9309, 2023. 3

    2023

  14. [22]

    Smpl: A skinned multi- person linear model

    Matthew Loper, Naureen Mahmood, Javier Romero, Gerard Pons-Moll, and Michael J Black. Smpl: A skinned multi- person linear model. In Seminal Graphics Papers: Pushing the Boundaries, Volume 2, pages 851–866. 2023. 1, 2, 3, 7

    2023

  15. [23]

    Decoupled weight decay regularization

    Ilya Loshchilov and Frank Hutter. Decoupled weight decay regularization. arXiv preprint arXiv:1711.05101, 2017. 6

    2017 arXiv

  16. [24]

    Dposer: Diffusion model as robust 3d human pose prior

    Junzhe Lu, Jing Lin, Hongkun Dou, Yulun Zhang, Yue Deng, and Haoqian Wang. Dposer: Diffusion model as robust 3d human pose prior. arXiv preprint arXiv:2312.05541, 2023. 6, 7, 8

    2023 arXiv

  17. [25]

    Diffusion probabilistic models for 3d point cloud generation

    Shitong Luo and Wei Hu. Diffusion probabilistic models for 3d point cloud generation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 2837–2845, 2021. 3

    2021

  18. [26]

    Controllable mesh generation through 9 sparse latent point diffusion models

    Zhaoyang Lyu, Jinyi Wang, Yuwei An, Ya Zhang, Dahua Lin, and Bo Dai. Controllable mesh generation through 9 sparse latent point diffusion models. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 271–280, 2023. 3

    2023

  19. [27]

    Amass: Archive of motion capture as surface shapes

    Naureen Mahmood, Nima Ghorbani, Nikolaus F Troje, Ger- ard Pons-Moll, and Michael J Black. Amass: Archive of motion capture as surface shapes. In Proceedings of the IEEE/CVF international conference on computer vision, pages 5442–5451, 2019. 6, 7

    2019

  20. [28]

    Difffacto: Controllable part-based 3d point cloud generation with cross diffusion

    George Kiyohiro Nakayama, Mikaela Angelina Uy, Jiahui Huang, Shi-Min Hu, Ke Li, and Leonidas Guibas. Difffacto: Controllable part-based 3d point cloud generation with cross diffusion. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 14257–14267...

    2023

  21. [29]

    The shape variational autoencoder: A deep generative model of part- segmented 3d objects

    Charlie Nash and Christopher KI Williams. The shape variational autoencoder: A deep generative model of part- segmented 3d objects. In Computer Graphics Forum, pages 1–12. Wiley Online Library, 2017. 3

    2017

  22. [30]

    Supr: A sparse unified part-based human representation

    Ahmed AA Osman, Timo Bolkart, Dimitrios Tzionas, and Michael J Black. Supr: A sparse unified part-based human representation. In European Conference on Computer Vi- sion, pages 568–585. Springer, 2022. 2

    2022

  23. [31]

    Expressive body capture: 3d hands, face, and body from a single image

    Georgios Pavlakos, Vasileios Choutas, Nima Ghorbani, Timo Bolkart, Ahmed AA Osman, Dimitrios Tzionas, and Michael J Black. Expressive body capture: 3d hands, face, and body from a single image. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognitio...

    2019

  24. [32]

    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,

  25. [33]

    Film: Visual reasoning with a general conditioning layer

    Ethan Perez, Florian Strub, Harm De Vries, Vincent Du- moulin, and Aaron Courville. Film: Visual reasoning with a general conditioning layer. In Proceedings of the AAAI con- ference on artificial intelligence, 2018. 5

    2018

  26. [34]

    Dreamfusion: Text-to-3d using 2d diffusion

    Ben Poole, Ajay Jain, Jonathan T Barron, and Ben Milden- hall. Dreamfusion: Text-to-3d using 2d diffusion. arXiv preprint arXiv:2209.14988, 2022. 3

    2022 arXiv

  27. [35]

    Diffusion autoen- coders: Toward a meaningful and decodable representation

    Konpat Preechakul, Nattanat Chatthee, Suttisak Wizad- wongsa, and Supasorn Suwajanakorn. Diffusion autoen- coders: Toward a meaningful and decodable representation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 10619–10629, 2022. 2

    2022

  28. [36]

    Pointnet: Deep learning on point sets for 3d classification and segmentation

    Charles R Qi, Hao Su, Kaichun Mo, and Leonidas J Guibas. Pointnet: Deep learning on point sets for 3d classification and segmentation. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 652–660,

  29. [37]

    Generating 3d faces using convolutional mesh autoencoders

    Anurag Ranjan, Timo Bolkart, Soubhik Sanyal, and Michael J Black. Generating 3d faces using convolutional mesh autoencoders. In Proceedings of the European confer- ence on computer vision (ECCV), pages 704–720, 2018. 7

    2018

  30. [38]

    The caesar project: a 3-d surface anthropometry survey

    Kathleen M Robinette, Hans Daanen, and Eric Paquet. The caesar project: a 3-d surface anthropometry survey. In Sec- ond international conference on 3-D digital imaging and modeling (cat. No. PR00062) , pages 380–386. IEEE, 1999. 6

    1999

  31. [39]

    High-resolution image synthesis with latent diffusion models

    Robin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser, and Bj ¨orn Ommer. High-resolution image synthesis with latent diffusion models. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 10684–10695, 2022. 2

    2022

  32. [40]

    Em- bodied hands: Modeling and capturing hands and bodies to- gether

    Javier Romero, Dimitrios Tzionas, and Michael J Black. Em- bodied hands: Modeling and capturing hands and bodies to- gether. arXiv preprint arXiv:2201.02610, 2022. 7

    2022 arXiv

  33. [41]

    Score-based generative modeling through stochastic differential equa- tions

    Yang Song, Jascha Sohl-Dickstein, Diederik P Kingma, Ab- hishek Kumar, Stefano Ermon, and Ben Poole. Score-based generative modeling through stochastic differential equa- tions. arXiv preprint arXiv:2011.13456, 2020. 2

    2011 arXiv

  34. [42]

    Applications of pose estimation in human health and performance across the lifespan

    Jan Stenum, Kendra M Cherry-Allen, Connor O Pyles, Rachel D Reetzke, Michael F Vignos, and Ryan T Roem- mich. Applications of pose estimation in human health and performance across the lifespan. Sensors, 21(21):7315,

  35. [43]

    Learning semantic- aware disentangled representation for flexible 3d human body editing

    Xiaokun Sun, Qiao Feng, Xiongzheng Li, Jinsong Zhang, Yu-Kun Lai, Jingyu Yang, and Kun Li. Learning semantic- aware disentangled representation for flexible 3d human body editing. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 16985...

    2023

  36. [44]

    Variational autoen- coder with implicit optimal priors

    Hiroshi Takahashi, Tomoharu Iwata, Yuki Yamanaka, Masanori Yamada, and Satoshi Yagi. Variational autoen- coder with implicit optimal priors. In Proceedings of the AAAI Conference on Artificial Intelligence , pages 5066– 5073, 2019. 5

    2019

  37. [45]

    Make-it-3d: High-fidelity 3d creation from a single image with diffusion prior

    Junshu Tang, Tengfei Wang, Bo Zhang, Ting Zhang, Ran Yi, Lizhuang Ma, and Dong Chen. Make-it-3d: High-fidelity 3d creation from a single image with diffusion prior. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 22819–22829, 2023. 3

    2023

  38. [46]

    Pose-ndf: Modeling human pose manifolds with neural distance fields

    Garvita Tiwari, Dimitrije Anti ´c, Jan Eric Lenssen, Nikolaos Sarafianos, Tony Tung, and Gerard Pons-Moll. Pose-ndf: Modeling human pose manifolds with neural distance fields. In European Conference on Computer Vision , pages 572–

  39. [47]

    Demea: Deep mesh au- toencoders for non-rigidly deforming objects

    Edgar Tretschk, Ayush Tewari, Michael Zollh¨ofer, Vladislav Golyanik, and Christian Theobalt. Demea: Deep mesh au- toencoders for non-rigidly deforming objects. In Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part IV 16 , p...

    2020

  40. [48]

    Nvae: A deep hierarchical vari- ational autoencoder

    Arash Vahdat and Jan Kautz. Nvae: A deep hierarchical vari- ational autoencoder. Advances in neural information pro- cessing systems, 33:19667–19679, 2020. 5

    2020

  41. [49]

    Score-based generative modeling in latent space

    Arash Vahdat, Karsten Kreis, and Jan Kautz. Score-based generative modeling in latent space. Advances in neural in- formation processing systems, 34:11287–11302, 2021. 5

    2021

  42. [50]

    Lion: Latent point dif- fusion models for 3d shape generation

    Arash Vahdat, Francis Williams, Zan Gojcic, Or Litany, Sanja Fidler, Karsten Kreis, et al. Lion: Latent point dif- fusion models for 3d shape generation. Advances in Neural Information Processing Systems, 35:10021–10039, 2022. 3

    2022

  43. [51]

    Attention is all you need

    Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszko- reit, Llion Jones, Aidan N Gomez, Łukasz Kaiser, and Illia 10 Polosukhin. Attention is all you need. Advances in neural information processing systems, 30, 2017. 1

    2017

  44. [52]

    Ghum & ghuml: Generative 3d human shape and articulated pose models

    Hongyi Xu, Eduard Gabriel Bazavan, Andrei Zanfir, William T Freeman, Rahul Sukthankar, and Cristian Smin- chisescu. Ghum & ghuml: Generative 3d human shape and articulated pose models. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages...

    2020

  45. [53]

    Semantic parametric reshaping of human body models

    Yipin Yang, Yao Yu, Yu Zhou, Sidan Du, James Davis, and Ruigang Yang. Semantic parametric reshaping of human body models. In 2014 2nd International Conference on 3D Vision, pages 41–48. IEEE, 2014. 6

    2014

  46. [54]

    Unsupervised shape and pose disentanglement for 3d meshes

    Keyang Zhou, Bharat Lal Bhatnagar, and Gerard Pons- Moll. Unsupervised shape and pose disentanglement for 3d meshes. In Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceed- ings, Part XXII 16 , pages 341–357. Springer, 2020. 2, 5, 7 11

    2020

  47. [589]

    Springer, 2022. 6, 7, 8

    2022

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.