Pith. sign in

REVIEW 4 major objections 6 minor 1 cited by

Quaffure: Real-Time Quasi-Static Neural Hair Simulation

T0 review · 4 major / 6 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read A self-supervised neural network predicts quasi-static hair drapes in milliseconds, without any simulated training data.

desk verdict A genuinely fast self-supervised hair simulator whose own numbers undercut its 'comparable quality' claim. read the letter →

arxiv 2412.10061 v2 pith:LNYATIVR submitted 2024-12-13 cs.CV cs.GR

Tuur Stuyck , Gene Wei-Chin Lin , Egor Larionov , Hsiao-yu Chen , Aljaz Bozic , Nikolaos Sarafianos , Doug Roble This is my paper · ORCID
classification cs.CVcs.GR
keywords hairsimulationself-supervisedlearningquasi-staticCosseratrodcollisiondetectionreal-timeanimationneuraldeformationgrooming
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 attempts to prove that a neural network can replace physics-based hair simulation for quasi-static draping in real-time applications. The authors claim their model predicts physically plausible hair positions in a few milliseconds on consumer hardware, generalizing across body poses, body shapes, and at least ten distinct hairstyles. The key advantage is that training requires no pre-computed simulation data; instead, a physics-based loss guides the network directly. If correct, this removes a major bottleneck for real-time avatars, games, and telepresence, where hair simulation is typically too slow or requires expensive offline data generation.

What carries the argument

The central mechanism is a modified Cosserat elastic energy that uses only positions, not orientations, to keep training efficient. The strain measure is defined as $\tilde{\Gamma} = (x_{i+1} - x_i)/l_{\text{rest}} - d_3$, where $d_3$ is the unit director along the rigidly transformed rest edge, paired with a Hookean stretch term to maintain rest length. This formulation preserves curl and requires orders-of-magnitude less training time than full Cosserat rod optimization. Complementing it are an SPH-based self-collision potential with a smooth kernel and a body collision potential using signed distances, all back-propagated through a 2D convolutional decoder conditioned on a latent groom code and body parameters.

What would settle it

Take a trained model and a pose sequence that includes a fast shoulder rotation and a bent neck. Compute the quasi-static drape with a full Cosserat rod simulation that includes strand friction, then compare strand-by-strand positions. If the network's drapes deviate beyond a tolerance (e.g., more than a few millimeters in strand tip position or show strand-body intersections that the simulation resolves), the simplified energy is missing a load-bearing physical effect.

Watch

Extended reading notes

Core claim

The paper's central claim is that hair quasi-statics can be learned entirely through self-supervision, using a differentiable energy loss that encodes stretch, bending, gravity, body collision, and hair self-collision. The method splits the problem into a rigid pose-based groom transformation, which moves the hair with the head, and a learned deformation decoder, which adds pose- and shape-dependent correctives. The decoder is trained to satisfy the physics energy, and the result is a network that positions hair vertices near the quasi-static energy minimum, producing smooth, collision-aware drapes at fixed inference cost independent of strand count.

Load-bearing premise

The load-bearing premise is that the hand-specified physics energy—with its stiffness weights and the simplified Cosserat term—faithfully captures the quasi-static behavior of real hair, including friction and persistent contact; if that energy omits important effects, the network will confidently produce plausible-looking but physically wrong drapes.

Editorial extensions

If this is right

  • Real-time hair draping for avatars becomes feasible on commodity hardware, with a fixed inference cost that does not grow with strand count or collision complexity.
  • The method can scale to a thousand hair grooms in 0.3 seconds, enabling crowd or multi-character scenarios.
  • Eliminating simulated training data removes the need for expert-parameterized offline simulation and large storage, simplifying the production pipeline.
  • Because the loss is physics-based, the network produces temporally smooth pose-dependent results and can generalize to body shapes not seen in training.

Reading between the lines

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

  • The same self-supervised physics-loss approach could extend to dynamic hair by conditioning on velocities and time, moving beyond quasi-statics to full motion.
  • Conditioning the decoder on material stiffness parameters at inference (rather than fixing them) would give artists control without retraining, a natural next step.
  • The modified Cosserat energy with position-only strain might transfer to other rod-like deformable objects, such as cables or cloth seams, where speed and stability are critical.
  • The fidelity of the predicted drape depends on the physics energy capturing persistent friction and static balance; a hybrid that adds a lightweight friction term could close the gap with full simulation.
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 proposes Quaffure, a neural quasi-static hair simulator that predicts draped hair in a few milliseconds on commodity hardware. The method decomposes hair deformation into a rigid pose-based groom transformation plus a learned deformation decoder conditioned on a 16-d groom latent code, body shape parameters, and skeleton pose. Training is self-supervised with a physics-based loss composed of a modified Cosserat elastic potential, stretch, gravity, body collision, SPH-based self-collision, and pose regularization (Eqs. 1-14), avoiding precomputed simulation data. Experiments compare against Adam, L-BFGS, XPBD, and a re-implementation of GroomGen, reporting quantitative metrics on length preservation, body intersection, and orientation preservation, as well as timing results showing 2.86 ms per groom and 0.3 s for 1000 grooms. The central claims are that Quaffure is the first self-supervised neural approach to real-time quasi-static hair simulation, produces physically plausible drapes of quality comparable to optimization-based simulation, and generalizes across hairstyles, body shapes, and poses.

Significance. If the claims hold, the work is practically significant: it removes the need for expensive simulated training data, provides fixed-cost inference independent of strand count, and demonstrates generalization over multiple grooms, poses, and shapes. The paper is clearly written and the timing measurements are credible. Notable strengths include the explicit decomposition into rigid transformation and learned deformation, the optimization-friendly modified Cosserat formulation, and the ablation against a mass-spring baseline. However, the quantitative validation is substantially weakened by three issues: the evaluation metrics overlap the training losses, the reported numbers in Table 1 do not support the 'comparable metrics' caption, and the GroomGen comparison relies on an unvalidated re-implementation. These issues directly affect the paper's central physical-plausibility and state-of-the-art claims, so the significance is conditional on additional independent evaluation.

major comments (4)
  1. [§4.2, Table 1; §3.5, Eqs. (6), (7), (10)] The quantitative evaluation is partly circular. The three reported metrics—length preservation, body-intersection percentage, and orientation preservation—are essentially the training losses Lstretch (Eq. 7), Lbody collision (Eq. 10), and the modified Cosserat term (Eq. 6), respectively. Table 1 therefore mainly shows how well the network minimizes its own training objective, not whether the resulting drapes are physically plausible. This is load-bearing for the central claim. I recommend adding an independent validation: for example, compare against real captured hair under controlled poses, or against a physics simulator using a different energy model (e.g., a full Cosserat rod or an established commercial solver), or run a perceptual user study. Without such evidence, 'physically plausible' is not independently established.
  2. [Table 1 and its caption] The caption states that 'Our method displays comparable metrics to directly optimizing for the positions,' but the numbers do not support this. Ours reports length preservation 175.42 versus 103.53 (Adam) and 89.53 (L-BFGS), and orientation preservation 286.13 versus 76.15 and 70.22—roughly 1.7 to 4 times worse on the paper's own metrics. Body intersection is also slightly worse (0.26 versus 0.22). Either the claim must be revised to reflect the actual gap, or the authors should provide a principled argument (e.g., a perceptual threshold) for why these differences do not affect the practical quality claim.
  3. [§4.2, GroomGen baseline] The comparison to GroomGen is based on the authors' own re-implementation, since the original code is not public. The re-implementation is not validated against the original method, and its training protocol differs from GroomGen's (random neck rotations instead of gravity variations). This makes the 'better than GroomGen' conclusion unreliable. I would ask the authors to either obtain the original implementation, release and validate their re-implementation against the paper's reported behavior, or substantially soften the comparative claim and present the result as indicative rather than definitive.
  4. [§4.2, Table 1] All quantitative comparisons are reported as single numbers with no error bars, no multiple seeds, and no statistical significance testing. Since the training involves stochastic optimization and the metrics may vary across seeds and dataset splits, the claimed margins over GroomGen and the 'comparable' phrasing relative to Adam/L-BFGS need variance information. At minimum, report mean and standard deviation over at least three training runs or over multiple evaluation subsets.
minor comments (6)
  1. [§3.5, Eq. (6)] The unit director d3 is described as 'computed from the rigidly transformed groom,' but no explicit formula is given. Please clarify how d3 is obtained from the rigid transform and how it relates to the rest-shape director.
  2. [§3.5, Eq. (13)] The piecewise SPH kernel has a typo: the middle branch should be (2 - r/h)^3 rather than '2 - r/h^3'. Also, '2 h ≤ r' should be written with a multiplication sign or space for clarity.
  3. [Table 1] The table mixes 'Gravity Potentials' with the other metrics but gives no units or indication whether lower is better for that column. Please state units and add a directional arrow, or move the gravity potential to a separate table.
  4. [§4, Dataset and implementation details] The dataset description is vague: it says 'CT-groom data set complemented with additional grooms made by technical artists' but gives no counts, no train/test split, and no statement of whether the evaluated grooms are held out from training. This is important for assessing the generalization claims.
  5. [§3.5, Eq. (14)] The pose regularization term uses Npose reg continuous frames, but the value of Npose reg is never specified. Please report it in the implementation details.
  6. [Figure 9] The axes and tick labels in Figure 9 are difficult to read, and the text '12 5 102' appears to be a formatting artifact. Please redraw the figure with clear axis labels and legible font sizes.

Circularity Check

2 steps flagged · score 6.0 of 10

Quantitative physical-plausibility validation is partially circular: Table 1 metrics are the same terms minimized by the self-supervised loss, and all baselines optimize or simulate that same energy.

  1. self definitional [Section 4 'Quantitative Comparisons' and Section 3.5 Eqs. (6)-(11), especially Eqs. (6), (7), (10)]
    "To evaluate our approach we opted for three key metrics: i) the length preservation of the hair strands measured on the segment lengths, ii) the percentage of intersections between the hair and the body, and iii) the preservation of the hair shape measured from the segment orientations."

    The three quantitative metrics are exactly the quantities minimized by the training loss. Length preservation is the Lstretch target (l - lrest)^2 in Eq. (7); body-intersection percentage is the max(D - d(x), 0)^3 target in Eq. (10); orientation preservation is the LCosserat director-difference target in Eq. (6). Therefore Table 1 mostly reports how well the network minimized its own self-supervised objective. Because 'physically plausible' is defined by this hand-set energy, the quantitative validation cannot independently establish physical plausibility; it only shows consistency with the training loss.

  2. other [Section 4, 'XPBD - Physics-based Simulation']
    "In our comparison, we model hair using the same energy as described in Subsection 3.5."

    The XPBD baseline, as well as the Adam and L-BFGS baselines that directly minimize the loss energy, all use the same energy as the training objective. Consequently, comparisons to these baselines are comparisons against other minimizers of the same self-defined energy, not against external physical ground truth. This reinforces that the 'comparable quality to optimization-based simulation' claim reduces to consistency with the paper's own loss rather than independent verification against real hair behavior or a different simulation model.

full rationale

The core learned mapping is not circular: the network is trained with an explicit physics-based loss and is evaluated on held-out poses, grooms, and body shapes, so the speed, generalization, and temporal-smoothness claims have independent content. The stiffness parameters are hand-set rather than fitted to the target data, so this is not a parameter-fitting circularity. The circularity is confined to the physical-plausibility validation: the three quantitative metrics are precisely the terms of the training loss (stretch, body collision, orientation/Cosserat), and the optimization/simulation baselines minimize or simulate that same energy, so Table 1 cannot distinguish a network that learned genuine quasi-static hair physics from one that learned to minimize a hand-tuned energy. The paper's own Table 1 also contains an internal inconsistency: Ours reports length preservation 175.42 and orientation preservation 286.13 versus Adam 103.53/76.15 and L-BFGS 89.53/70.22, contradicting the caption's 'comparable metrics' claim; this is a correctness concern rather than a circularity. No load-bearing self-citation chain or imported uniqueness theorem was found. Score 6 reflects that the central quantitative claim is partially circular while the speed and generalization claims remain independent.

Assumptions & free parameters 7 free parameters · 5 assumptions · 0 invented entities

The method depends on the adequacy of its physics loss as a training signal and on the representation capacity of the decoder. The stiffness and collision constants are hand-set, not fitted to ground-truth data, so the model is not circular in the fitting sense, but the validation metrics overlap with the training loss.

free parameters (7)
  • kCosserat (Cosserat stiffness)
    Hand-tuned stiffness controlling shear and bending resistance in Eq. 6; value not reported.
  • kstretch (stretch stiffness)
    Hand-tuned penalty for segment length deviation from rest in Eq. 7.
  • kbc (body collision stiffness)
    Hand-tuned weight for body collision loss in Eq. 10.
  • ksc (self-collision stiffness)
    Hand-tuned weight for SPH self-collision loss in Eq. 11.
  • D (minimum body collision distance)
    Required distance between hair vertices and body surface in Eq. 10.
  • h (SPH smoothing length)
    Radius of the SPH kernel W in Eq. 13, sets self-collision interaction range.
  • kpr (pose regularization weight)
    Weight for temporal smoothness loss in Eq. 14.
assumptions (5)
  • domain assumption Quasi-static equilibrium: hair positions are determined by the minimizer of the energy potentials for each pose and shape; dynamic effects are neglected.
    The paper models quasi-static drapes only; title and Section 3.5. No validation against dynamic hair.
  • domain assumption The modified Cosserat energy with position-only offsets (Eq. 6) plus Hookean stretch (Eq. 7) approximates hair elastic behavior, especially for curly strands.
    Section 3.5, Eq. 6-8. The full Cosserat rod model was abandoned due to training time; the simplified version is used without comparison to real hair mechanics.
  • domain assumption A fixed global stiffness parameter set is sufficient for all hairstyles in the study (curly, straight, long, short).
    Section 4.4 says current results use a single set of physical parameters; this assumes one material model spans all grooms.
  • ad hoc to paper The 16-dimensional groom autoencoder latent code is sufficient to represent all training grooms and to condition the deformation decoder.
    Section 3.2; the autoencoder is trained on rest grooms, but the paper does not report reconstruction error or evaluate the latent space's completeness.
  • domain assumption The motion capture pose distribution used for training is representative of the target inference distribution.
    Section 4.1 limitations note that non-physical poses absent from training produce implausible results, implying this coverage assumption is load-bearing.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Quaffure: Real-Time Quasi-Static Neural Hair Simulation." pith.science (2026). https://pith.science/paper/LNYATIVR

@misc{pith2026241210061,
  author       = {Pith},
  title        = {Pith review of: Quaffure: Real-Time Quasi-Static Neural Hair Simulation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/LNYATIVR}},
  note         = {Machine review of arXiv:2412.10061}
}
read the original abstract

Realistic hair motion is crucial for high-quality avatars, but it is often limited by the computational resources available for real-time applications. To address this challenge, we propose a novel neural approach to predict physically plausible hair deformations that generalizes to various body poses, shapes, and hairstyles. Our model is trained using a self-supervised loss, eliminating the need for expensive data generation and storage. We demonstrate our method's effectiveness through numerous results across a wide range of pose and shape variations, showcasing its robust generalization capabilities and temporally smooth results. Our approach is highly suitable for real-time applications with an inference time of only a few milliseconds on consumer hardware and its ability to scale to predicting the drape of 1000 grooms in 0.3 seconds. Please see our project page here following https://tuurstuyck.github.io/quaffure/quaffure.html

Figures

Figures reproduced from arXiv: 2412.10061 by the authors.

Figure 1
Figure 1. We present Quaffure, a real-time quasi-static neural hair simulator, which produces naturally draped hair in only a few millisec￾onds on commodity hardware, taking the hairstyle, body shape and pose into account. Our method scales to predicting the drape of 1000 hair grooms in just 0.3 seconds. Quaffure is trained using a physics-based self-supervised loss, eliminating the need for simulated training data that is co… view at source ↗
Figure 2
Figure 2. Quaffure Overview: Our method takes a code as input, consisting of a latent code for the rest hair shape, body shape parameters, and full skeleton pose. The output is naturally draped hair produced as the sum of posed hair given the body pose and shape parameters, combined with learned corrections which are produced by the groom deformation decoder. We train our method in two stages: i) an autoencoder is trained on … view at source ↗
Figure 3
Figure 3. Pose & Shape-based Deformations: Example of posed groom (top) which accounts for rigid rotations and translations of the rest shape only and the combined posed groom with learned deformations (bottom), which accounts for physical effects such as strand material model, gravity and collisions. Despite body intersections incurred by the rigid transformation applied by the groom transformation module, our proposed netwo… view at source ↗
Figures from the paper (7 more)
Figure 4
Figure 4. Figure 4: Modeling Diverse Hairstyles: We showcase the versatility of our method in modeling the quasi-static behavior of a plethora of hairstyles and lengths for different body shapes and poses where all results are obtained using the same settings without any manual parameter …
Figure 5
Figure 5. Figure 5: Resolving Collisions: Our method is conditioned on the body shape parameters, enabling it to efficiently resolve collisions with body shape variations. Here we show a static pose and groom under quasi-statically draped under varying body shapes. tions between the hair …
Figure 8
Figure 8. Figure 8: Comparisons with GroomGen: We compare our re￾sults (bottom) to those obtained with the neural simulator proposed in GroomGen [51] (top). The inset figure highlights that Groom￾Gen produces results where strands intersect with the head geome￾try. In contrast, our method…
Figure 6
Figure 6. Figure 6: Temporal Stability: To demonstrate how predicted re￾sults are smoothly varying with pose changes, we gradually mod￾ify the neck rotation. Starting from looking to the right to looking left. Note how the results are smoothly varying where every pose produces a natural d…
Figure 7
Figure 7. Figure 7: Qualitative Comparisons: Our method produces visu￾ally comparable results to directly minimizing the energies or to quasi-static simulation with XPBD while being orders of magni￾tude faster at inference by memorizing body pose and shape de￾pendent groom deformations. p…
Figure 9
Figure 9. Figure 9: Performance & Scaling: Our method takes less than 3 ms on average to predict a single groom and the time complexity scales linearly with the number of grooms. The method is able to predict a thousand hairstyles in 0.3 seconds [PITH_FULL_IMAGE:figures/full_fig_p008_9.png]
Figure 10
Figure 10. Figure 10: Ablation Study: We perform an ablation of our pro￾posed optimization-friendly Cosserat model against the commonly used mass-spring energy for modeling hair. Our model excels at maintaining the desired hairstyle whereas the mass-spring model struggles with maintaining …

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. HairFormer: Transformer-Based Dynamic Neural Hair Simulation

    cs.GR 2025-07 conditional novelty 6.0 of 10

    A transformer-based two-stage network predicts static hair drapes and dynamic hair motion for arbitrary hairstyles and body poses in real time, trained with physics-inspired losses rather than pre-simulated data.

Reference graph

Works this paper leans on

52 extracted references · 46 canonical work pages · cited by 1 Pith paper

  1. [1]

    Animating hair with loosely connected particles

    Yosuke Bando, Bing-Yu Chen, and Tomoyuki Nishita. Animating hair with loosely connected particles. In Computer Graphics F orum , pages 411–418. Wiley Online Library, 2003. 2, 5

    work page 2003

  2. [2]

    Discrete elastic rods

    Mikl ´os Bergou, Max Wardetzky, Stephen Robinson, Basile Audoly, and Eitan Grinspun. Discrete elastic rods. In ACM SIGGRAPH 2008 Papers . Association for Computing Machinery, 2008. 2, 3

    work page 2008

  3. [3]

    Neural cloth simulation

    Hugo Bertiche, Meysam Madadi, and Sergio Escalera. Neural cloth simulation. ACM Transactions on Graph- ics (TOG), 41(6):1–14, 2022. 3

    work page 2022

  4. [4]

    Real-time hair rendering with hair meshes

    Gaurav Bhokare, Eisen Montalvo, Elie Diaz, and Cem Yuksel. Real-time hair rendering with hair meshes. In ACM SIGGRAPH 2024 Conference Papers , pages 1–10, 2024. 3

    2024

  5. [5]

    Projective dynam- ics: Fusing constraint projections for fast simulation

    Sofien Bouaziz, Sebastian Martin, Tiantian Liu, Ladislav Kavan, and Mark Pauly. Projective dynam- ics: Fusing constraint projections for fast simulation. In Seminal Graphics Papers: Pushing the Boundaries, V olume 2, pages 787–797. Association for Computing Machinery, 2023. 2

    work page 2023

  6. [6]

    A re- duced model for interactive hairs

    Menglei Chai, Changxi Zheng, and Kun Zhou. A re- duced model for interactive hairs. ACM Transactions on Graphics (TOG), 33(4):1–11, 2014. 2, 3

    work page 2014

  7. [7]

    Interactive hair simulation on the gpu using admm

    Gilles Daviet. Interactive hair simulation on the gpu using admm. In ACM SIGGRAPH 2023 Conference Proceedings, pages 1–11, 2023. 1, 2

    work page 2023

  8. [8]

    Drapenet: Garment generation and self-supervised draping

    Luca De Luigi, Ren Li, Benoit Guillard, Mathieu Salz- mann, and Pascal Fua. Drapenet: Garment generation and self-supervised draping. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pat- tern Recognition, pages 1451–1460, 2023. 3

    work page 2023

Show all 52 references
  1. [9]

    Latent-space dy- namics for reduced deformable simulation

    Lawson Fulton, Vismay Modi, David Duvenaud, David IW Levin, and Alec Jacobson. Latent-space dy- namics for reduced deformable simulation. In Com- puter graphics forum , pages 379–391. Wiley Online Library, 2019. 3

    2019

  2. [10]

    Hood: Hierarchical graphs for generalized mod- elling of clothing dynamics

    Artur Grigorev, Michael J Black, and Otmar Hilliges. Hood: Hierarchical graphs for generalized mod- elling of clothing dynamics. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pat- tern Recognition, pages 16965–16974, 2023. 3

    2023

  3. [11]

    Multi-linear data-driven dy- namic hair model with efficient hair-body collision handling

    Peng Guan, Leonid Sigal, Valeria Reznitskaya, and Jessica K Hodgins. Multi-linear data-driven dy- namic hair model with efficient hair-body collision handling. In Proceedings of the 11th ACM SIG- GRAPH/Eurographics conference on Computer Ani- mation, pages 295–304, 2012. 1, 3

    2012

  4. [12]

    Perm: A parametric rep- resentation for multi-style 3d hair modeling

    Chengan He, Xin Sun, Zhixin Shu, Fujun Luan, S ¨oren Pirk, Jorge Alejandro Amador Herrera, Dominik L Michels, Tuanfeng Y Wang, Meng Zhang, Holly Rushmeier, and Yi Zhou. Perm: A parametric rep- resentation for multi-style 3d hair modeling. arXiv preprint arXiv:2407.19451, 2024. 2

    2024 arXiv

  5. [13]

    Real-time physi- cally guided hair interpolation

    Jerry Hsu, Tongtong Wang, Zherong Pan, Xifeng Gao, Cem Yuksel, and Kui Wu. Real-time physi- cally guided hair interpolation. ACM Transactions on Graphics (TOG), 43(4):1–11, 2024. 2, 3, 6

    2024

  6. [14]

    Simulation-ready hair capture

    Liwen Hu, Derek Bradley, Hao Li, and Thabo Beeler. Simulation-ready hair capture. In Computer Graphics F orum, pages 281–294. Wiley Online Library, 2017. 3

    2017

  7. [15]

    Towards realtime: A hybrid physics-based method for hair animation on gpu

    Li Huang, Fan Yang, Chendi Wei, Yu Ju Chen, Chun Yuan, and Ming Gao. Towards realtime: A hybrid physics-based method for hair animation on gpu. Pro- ceedings of the ACM on Computer Graphics and In- teractive Techniques, 6(3):1–18, 2023. 1

    2023

  8. [16]

    Neuralclothsim: Neural deformation fields meet the thin shell theory

    Navami Kairanda, Marc Habermann, Christian Theobalt, and Vladislav Golyanik. Neuralclothsim: Neural deformation fields meet the thin shell theory. In Neural Information Processing Systems (NeurIPS) ,

  9. [17]

    Kaufman, Rasmus Tamstorf, Breannan Smith, Jean-Marie Aubry, and Eitan Grinspun

    Danny M. Kaufman, Rasmus Tamstorf, Breannan Smith, Jean-Marie Aubry, and Eitan Grinspun. Adap- tive nonlinearity for collisions in complex rod assem- blies. ACM Trans. Graph., 33(4), 2014. 3

    2014

  10. [18]

    3d gaussian splatting for real-time radiance field rendering

    Bernhard Kerbl, Georgios Kopanas, Thomas Leimk¨uhler, and George Drettakis. 3d gaussian splatting for real-time radiance field rendering. ACM Transactions on Graphics, 42(4), 2023. 2

    2023

  11. [19]

    Real-time an- imation of human hair modeled in strips

    Chuan Koon Koh and Zhiyong Huang. Real-time an- imation of human hair modeled in strips. In Com- puter Animation and Simulation 2000: Proceedings of the Eurographics Workshop in Interlaken, Switzer- land, August 21–22, 2000 , pages 101–110. Springer,

  12. [20]

    A simple physics model to animate human hair modeled in 2d strips in real time

    Chuan Koon Koh and Zhiyong Huang. A simple physics model to animate human hair modeled in 2d strips in real time. In Computer Animation and Sim- ulation 2001: Proceedings of the Eurographics Work- shop in Manchester , UK, September 2–3, 2001, pages 127–138. Springer, 2001. 3

    2001

  13. [21]

    Kugelstadt and E

    T. Kugelstadt and E. Sch ¨omer. Position and orienta- tion based cosserat rods. In Proceedings of the ACM SIGGRAPH/Eurographics Symposium on Computer Animation, page 169–178, Goslar, DEU, 2016. Euro- graphics Association. 2, 3, 4

    2016

  14. [22]

    Quasi-newton methods for real-time simulation of hy- perelastic materials

    Tiantian Liu, Sofien Bouaziz, and Ladislav Kavan. Quasi-newton methods for real-time simulation of hy- perelastic materials. ACM Transactions on Graphics (TOG), 36(3):23, 2017. 7

    2017

  15. [23]

    Gaussianhair: Hair modeling and rendering with light-aware gaussians

    Haimin Luo, Min Ouyang, Zijun Zhao, Suyi Jiang, Longwen Zhang, Qixuan Zhang, Wei Yang, Lan Xu, and Jingyi Yu. Gaussianhair: Hair modeling and rendering with light-aware gaussians. arXiv preprint arXiv:2402.10483, 2024. 2

    2024 arXiv

  16. [24]

    Real-time hair simulation with neural interpolation

    Qing Lyu, Menglei Chai, Xiang Chen, and Kun Zhou. Real-time hair simulation with neural interpolation. IEEE Transactions on Visualization and Computer Graphics, 28(4):1894–1905, 2020. 2, 3

    1905

  17. [25]

    Xpbd: position-based simulation of com- pliant constrained dynamics

    Miles Macklin, Matthias M ¨uller, and Nuttapong Chentanez. Xpbd: position-based simulation of com- pliant constrained dynamics. In Proceedings of the 9th International Conference on Motion in Games , pages 49–54, 2016. 2, 6, 7

    2016

  18. [26]

    Detail preserving continuum simulation of straight hair

    Aleka McAdams, Andrew Selle, Kelly Ward, Efty- chios Sifakis, and Joseph Teran. Detail preserving continuum simulation of straight hair. ACM Trans- actions on Graphics (ToG), 28(3):1–6, 2009. 2, 3, 5

    2009

  19. [27]

    Position based dynamics

    Matthias M ¨uller, Bruno Heidelberger, Marcus Hennix, and John Ratcliff. Position based dynamics. Journal of Visual Communication and Image Representation , 18(2):109–118, 2007. 2

    2007

  20. [28]

    Au- tomatic differentiation in pytorch, 2017

    Adam Paszke, Sam Gross, Soumith Chintala, Gregory Chanan, Edward Yang, Zachary DeVito, Zeming Lin, Alban Desmaison, Luca Antiga, and Adam Lerer. Au- tomatic differentiation in pytorch, 2017. 6

    2017

  21. [29]

    V ol- umetric methods for simulation and rendering of hair

    Lena Petrovic, Mark Henne, and John Anderson. V ol- umetric methods for simulation and rendering of hair. Pixar Animation Studios, 2(4):1–6, 2005. 2, 5

    2005

  22. [30]

    Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equa- tions

    Maziar Raissi, Paris Perdikaris, and George E Karni- adakis. Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equa- tions. Journal of Computational physics , 378:686– 707, 2019. 3

    2019

  23. [31]

    Neural strands: Learning hair geometry and appearance from multi-view images

    Radu Alexandru Rosu, Shunsuke Saito, Ziyan Wang, Chenglei Wu, Sven Behnke, and Giljoo Nam. Neural strands: Learning hair geometry and appearance from multi-view images. ECCV, 2022. 2

    2022

  24. [32]

    Snug: Self-supervised neural dynamic garments

    Igor Santesteban, Miguel A Otaduy, and Dan Casas. Snug: Self-supervised neural dynamic garments. In Proceedings of the IEEE/CVF Conference on Com- puter Vision and Pattern Recognition , pages 8140– 8150, 2022. 3

    2022

  25. [33]

    A mass spring model for hair simulation

    Andrew Selle, Michael Lentine, and Ronald Fedkiw. A mass spring model for hair simulation. InACM SIG- GRAPH 2008 papers, pages 1–11, 2008. 2, 3, 8

    2008

  26. [34]

    Data-free learning of reduced-order kine- matics

    Nicholas Sharp, Cristian Romero, Alec Jacobson, Eti- enne V ouga, Paul Kry, David IW Levin, and Justin Solomon. Data-free learning of reduced-order kine- matics. In ACM SIGGRAPH 2023 Conference Pro- ceedings, pages 1–9, 2023. 3

    2023

  27. [35]

    Ct2hair: High-fidelity 3d hair mod- eling using computed tomography

    Yuefan Shen, Shunsuke Saito, Ziyan Wang, Olivier Maury, Chenglei Wu, Jessica Hodgins, Youyi Zheng, and Giljoo Nam. Ct2hair: High-fidelity 3d hair mod- eling using computed tomography. ACM Transactions on Graphics (TOG), 42(4):1–13, 2023. 2, 5

    2023

  28. [36]

    Neural haircut: Prior-guided strand-based hair recon- struction

    Vanessa Sklyarova, Jenya Chelishev, Andreea Dogaru, Igor Medvedev, Victor Lempitsky, and Egor Zakharov. Neural haircut: Prior-guided strand-based hair recon- struction. In Proceedings of the IEEE/CVF Interna- tional Conference on Computer Vision , pages 19762– 19773, 2023. 2

    2023

  29. [37]

    Haar: Text- conditioned generative model of 3d strand-based hu- man hairstyles

    Vanessa Sklyarova, Egor Zakharov, Otmar Hilliges, Michael J Black, and Justus Thies. Haar: Text- conditioned generative model of 3d strand-based hu- man hairstyles. arXiv preprint arXiv:2312.11666 ,

    arXiv

  30. [38]

    Cosserat rods with projective dynamics

    Carlota Soler, Tobias Martin, and Olga Sorkine- Hornung. Cosserat rods with projective dynamics. Computer Graphics F orum, 37(8):137–147, 2018. 2, 3

    2018

  31. [39]

    Neuwigs: A neural dynamic model for volumetric hair capture and animation

    Ziyan Wang, Giljoo Nam, Tuur Stuyck, Stephen Lom- bardi, Chen Cao, Jason Saragih, Michael Zollhoefer, Jessica Hodgins, and Christoph Lassner. Neuwigs: A neural dynamic model for volumetric hair capture and animation. arXiv preprint arXiv:2212.00613, 2022. 1, 2

    2022 arXiv

  32. [40]

    Hvh: Learning a hybrid neural vol- umetric representation for dynamic hair performance capture

    Ziyan Wang, Giljoo Nam, Tuur Stuyck, Stephen Lombardi, Michael Zollh ¨ofer, Jessica Hodgins, and Christoph Lassner. Hvh: Learning a hybrid neural vol- umetric representation for dynamic hair performance capture. In Proceedings of the IEEE/CVF Conference on Computer Vision and ...

    2022

  33. [41]

    Haomiao Wu, Alvin Shi, J Parr, A. M. Darke, and Theodore Kim. Lifted curls: A model for tightly coiled hair simulation. In Proceedings of the ACM SIGGRAPH/Eurographics Symposium on Computer Animation. Eurographics Association, 2023. 3

    2023

  34. [42]

    Haomiao Wu, Alvin Shi, A. M. Darke, and Theodore Kim. Curly-cue: Geometric methods for highly coiled hair. In ACM SIGGRAPH Asia 2024 Conference Proceedings. Association for Computing Machinery,

    2024

  35. [43]

    Real-time hair mesh simu- lation

    Kui Wu and Cem Yuksel. Real-time hair mesh simu- lation. In Proceedings of the 20th ACM SIGGRAPH Symposium on Interactive 3D Graphics and Games , page 59–64, New York, NY , USA, 2016. Association for Computing Machinery. 1, 2

    2016

  36. [44]

    Neuralhdhair: Automatic high-fidelity hair modeling from a single image using implicit neural representations

    Keyu Wu, Yifan Ye, Lingchen Yang, Hongbo Fu, Kun Zhou, and Youyi Zheng. Neuralhdhair: Automatic high-fidelity hair modeling from a single image using implicit neural representations. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pat- tern Recognition, pages ...

    2022

  37. [45]

    Gaussian head avatar: Ultra high-fidelity head avatar via dynamic gaussians

    Yuelang Xu, Benwang Chen, Zhe Li, Hongwen Zhang, Lizhen Wang, Zerong Zheng, and Yebin Liu. Gaussian head avatar: Ultra high-fidelity head avatar via dynamic gaussians. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pat- tern Recognition (CVPR), 2024. 2

    2024

  38. [46]

    Dynamic hair modeling from monocular videos using deep neural networks

    Lingchen Yang, Zefeng Shi, Youyi Zheng, and Kun Zhou. Dynamic hair modeling from monocular videos using deep neural networks. ACM Transactions on Graphics (TOG), 38(6):1–12, 2019. 2

    2019

  39. [47]

    Hair meshes

    Cem Yuksel, Scott Schaefer, and John Keyser. Hair meshes. ACM Transactions on Graphics (TOG) , 28 (5):1–7, 2009. 3

    2009

  40. [48]

    Human hair reconstruction with strand-aligned 3d gaussians

    Egor Zakharov, Vanessa Sklyarova, Michael J Black, Giljoo Nam, Justus Thies, and Otmar Hilliges. Human hair reconstruction with strand-aligned 3d gaussians. ArXiv, 2024. 2

    2024

  41. [49]

    Ntopo: Mesh-free topology optimiza- tion using implicit neural representations

    Jonas Zehnder, Yue Li, Stelian Coros, and Bernhard Thomaszewski. Ntopo: Mesh-free topology optimiza- tion using implicit neural representations. Advances in Neural Information Processing Systems, 34:10368– 10381, 2021. 3

    2021

  42. [50]

    Hairnet: Single- view hair reconstruction using convolutional neural networks

    Yi Zhou, Liwen Hu, Jun Xing, Weikai Chen, Han- Wei Kung, Xin Tong, and Hao Li. Hairnet: Single- view hair reconstruction using convolutional neural networks. In Proceedings of the European Conference on Computer Vision (ECCV), pages 235–251, 2018. 2

    2018

  43. [51]

    Groomgen: A high- quality generative hair model using hierarchical la- tent representations

    Yuxiao Zhou, Menglei Chai, Alessandro Pepe, Markus Gross, and Thabo Beeler. Groomgen: A high- quality generative hair model using hierarchical la- tent representations. ACM Transactions on Graphics (TOG), 42(6):1–16, 2023. 2, 3, 6, 7

    2023

  44. [52]

    Groomcap: High-fidelity prior-free hair capture

    Yuxiao Zhou, Menglei Chai, Daoye Wang, Sebastian Winberg, Erroll Wood, Kripasindhu Sarkar, Markus Gross, and Thabo Beeler. Groomcap: High-fidelity prior-free hair capture. ACM Transactions on Graph- ics, 2024. 2

    2024

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.