Pith. sign in

REVIEW 3 major objections 6 minor 53 references

DeepSketchHair: Deep Sketch-based 3D Hair Modeling

T0 review · 3 major / 6 minor · reviewed 2026-08-14 · deepseek-v4-flash

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

Pith's one-line read Three networks turn 2D hair sketches into editable 3D hairstyles

desk verdict First deep-learning sketch-to-3D-hair pipeline, coherent and worth refereeing, but the evaluation leans on machine-traced training sketches and no released code, so the headline claim of matching arbitrary sketches is not fully nailed. read the letter →

arxiv 1908.07198 v1 pith:UV6MC5F5 submitted 2019-08-20 cs.GR

Yuefan Shen , Changgeng Zhang , Hongbo Fu , Kun Zhou , Youyi Zheng This is my paper · ORCID
classification cs.GR
keywords sketch-based3Dmodelinghairgenerativeadversarialnetworksorientationfieldstrand-levelsynthesismulti-vieweditingvoxelrepresentationinteractive
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

The paper tries to establish that a casual 2D sketch, consisting of a hair contour and a few direction strokes, can be automatically converted into a full strand-level 3D hairstyle that matches the sketch both globally and locally. It claims to be the first deep learning system for sketch-based 3D hair modeling, supporting both single-view creation and multi-view editing. The paper argues that a chain of intermediate representations, rather than direct sketch-to-3D regression, resolves the inherent sparsity and ambiguity of freehand sketches. A sympathetic reader should care because hair is notoriously difficult to model in 3D, existing sketch tools are labor-intensive or produce low-quality wisps, and the system promises realistic results in minutes with minimal user effort.

What carries the argument

The load-bearing object is the dense 2D orientation field, an intermediate map that gives every pixel in the hair region a direction vector for the strand passing through it. This representation bridges the gap between sparse, ambiguous sketch strokes and the full 3D volume, so the network never has to jump directly from a few strokes to 3D structure. The rest of the machinery consists of the three conditional GANs built on WGAN-GP, together with projection, Laplacian, and preservation losses that keep the synthesized 3D orientation field consistent with the input view and with previously generated hair during multi-view editing.

What would settle it

Collect freehand sketches from users who are told to draw freely, with gaps, overlapping strokes, and inconsistent directions, then feed those sketches into the trained pipeline and compare the grown strands against the users' intended hairstyles; high orientation error or consistent user rejection would refute the generalization claim. A cheaper check is to compare S2ONet's orientation output on real user sketches against the orientation map obtained by tracing that same output and re-inputting it.

Watch

Extended reading notes

Core claim

The central claim is that sparse, ambiguous 2D sketches can be mapped to dense strand-level 3D hair models through a learned pipeline of three generative adversarial networks. S2ONet fills the sketch and mask into a dense 2D orientation field, where every pixel stores the direction of the visible hair strand passing through it. O2VNet lifts that 2D orientation field, aided by the bust depth map, into a 128x128x96 3D orientation field from which strands are grown. V2VNet updates the 3D orientation field when the user draws additional sketches in new views, while preserving the parts of the volume not touched by the new input. The networks are trained on 653 strand-level hairstyles, using projection and Laplacian losses to tie the generated volume to the input view and to spread local constraints throughout the volume.

Load-bearing premise

The networks are trained only on synthetic sketches traced from the very orientation maps they must predict, so the claim that real freehand sketches produce desired hairstyles rests on the assumption that casual user drawings closely resemble those clean, perfectly consistent traces.

Editorial extensions

If this is right

  • A single hair contour plus a handful of direction strokes can produce a complete strand-level 3D hairstyle, cutting modeling time to minutes for novice users.
  • The same pipeline can ingest portrait images or cartoon references by automatically extracting masks and orientation maps, and then lets users refine the result by sketching in other views.
  • Using an intermediate dense 2D orientation field avoids the ambiguity of direct sketch-to-3D learning and preserves local detail from the input view.
  • Multi-view editing with V2VNet changes only the relevant region while keeping the rest of the hairstyle intact, enabling iterative refinement of a 3D hair model.
  • The approach is bounded by voxel resolution and training data, so fine details such as Afro textures and unseen shapes like polystrip styles are not faithfully reproduced.

Reading between the lines

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

  • Going beyond the paper: if the synthetic-to-real gap is closed by augmenting training with real traced sketches, the same two-stage design could apply to other strand-like structures such as fur, grass, or braided patterns.
  • Going beyond the paper: the sketch-to-dense-2D-field-to-3D-volume architecture suggests that for any sparse and ambiguous input, predicting an intermediate dense field aligned with the output view may be more effective than direct regression to the 3D target.
  • Going beyond the paper: the V2VNet strategy of preserving invisible voxels while updating visible ones could be reused for other local 3D editing tasks, where a targeted change should not disturb the rest of a pre-existing model.
  • Going beyond the paper: a direct measurement of how far real freehand sketches are from the clean traced training distribution, followed by stroke-aware data augmentation, would be a testable next step toward robust generalization.
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

3 major / 6 minor

Summary. The paper presents DeepSketchHair, an interactive system that converts 2D hair sketches (a contour plus a few direction strokes) into strand-level 3D hair models. The pipeline comprises three GAN-based networks: S2ONet maps the input sketch and mask to a dense 2D orientation field; O2VNet maps this field together with a bust depth map to a 3D volumetric orientation field; and V2VNet updates the volume from additional sketches drawn in rotated views. The networks are trained on synthetic data generated from 653 3D hair models, and the paper includes a qualitative gallery, a five-user pilot study, comparisons with prior image-based and sketch-based hair modeling methods, and ablations of the main components. The central claim is that the system generates 3D hair models that match the input sketch both globally and locally, and that it is the first deep-learning technique for sketch-based strand-level hair modeling.

Significance. If the central claim is accepted, this would be a useful contribution as the first deep-learning pipeline for sketch-based strand-level hair modeling and interactive multi-view editing. The architecture choices are clearly described and mostly well motivated, particularly the intermediate dense 2D orientation field, the projection and Laplacian losses for O2VNet, and the dedicated V2VNet for multi-view updates. The qualitative results and comparisons are visually compelling, and the S2ONet ablation is directionally consistent with the method's design. However, the paper currently rests on an untested synthetic-to-real sketch transfer assumption and on an evaluation that is too anecdotal and underpowered to fully support the claim of matching arbitrary freehand sketches. The absence of released code or data also limits independent verification, though that alone is not a reason to reject.

major comments (3)
  1. [§4.3, 2D Sketch Map generation] The tracing procedure as written is internally inconsistent. The text states that a neighbor is marked as a candidate when the dot product between p and pn is below the threshold ϱ = 0.5, and then "we choose the most matching pixel among the candidate pixels as the next iterative pixel by finding the minimum dot product." If taken literally, this follows the least-aligned neighbor rather than the most aligned one; it is presumably meant to be "above" and "maximum." Since the synthetic sketch maps are the only training inputs for S2ONet and the paper's generalization claim depends on their resemblance to real freehand strokes, this ambiguity should be corrected and the resulting maps verified. If the literal reading is intended, the maps are not clean orientation-consistent traces, and the connection between training and test inputs becomes even weaker.
  2. [§7.1, Pilot Study] The pilot study does not provide quantitative evidence that DeepSketchHair matches real freehand sketches. It uses five novice users, reports only timing and anecdotal "positive feedback," and contains no objective measure of whether the generated hair matches user intent (for example, agreement between the S2ONet-predicted orientation and the user's intended direction, task completion rate, or a comparison against a baseline). Because S2ONet is trained exclusively on machine-traced sketch maps generated from ground-truth orientation fields (§4.3), the system's central claim of handling casual user sketches rests on an unmeasured distribution-transfer assumption. The authors should add a real-sketch evaluation, ideally with multiple users and objective error metrics, before the main claim can be accepted.
  3. [§7.2 and §7.3, Quantitative comparisons and ablations] The quantitative evidence is too thin to support the comparative and ablative claims. Table 3 reports MSE values without error bars, test-set sizes, or statistical significance tests; Fig. 15 reports MSE numbers for O2VNet variants without any details on the test sample or variance; and §7.2 compares against prior work mostly qualitatively. Since the paper's conclusion emphasizes "outstanding performance," at least the main ablation claims need standard deviations, number of test examples, and a clear statement of how the test set is split from the 653-model dataset.
minor comments (6)
  1. [§4.1, Eq. (1)] The sets m and n used for content and style layers in Eq. (1) are not defined before the equation; the sentence describing m={0,2} and n={0,1,2,3,4} should be moved before Eq. (1) or integrated into its explanation.
  2. [§4.1, Network architecture] The phrase "we enclasp them into a 2D tensor" is unusual wording; consider replacing it with "concatenate" or "combine."
  3. [§5.3, Hair Synthesis] The sentence "our hair strands are more continuous and venerable in noise regions" appears to be a typo; likely "robust to noise" or "more stable in noise regions" was intended.
  4. [§4.2, Network Architecture] The sentence "We share the same discriminator architecture with zhang and zheng [9]" should capitalize the author names as "Zhang and Zheng."
  5. [§4.1 and §4.2, Notation] The ground-truth 2D orientation is denoted ~Λ in Eqs. (2) and (3), but the text surrounding Eq. (1) sometimes uses Λ for the target; please define both symbols explicitly and use them consistently.
  6. [Table 1] A short caption or a note explaining the meaning of the "views" and "strokes" columns would help readers interpret the interaction statistics.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the system is a supervised three-network pipeline with independent ground-truth supervision; synthetic sketch generation is a standard sparse-to-dense training construction, not a circular reduction.

full rationale

DeepSketchHair is an empirical machine-learning system rather than a formal derivation, and its claimed results are evaluated against external benchmarks and ground-truth 3D hair models. Each network is trained with supervised losses: S2ONet regresses a dense 2D orientation field from sparse sketch and mask inputs and is directly compared to the ground-truth dense orientation map (Eq. 2 with per-pixel loss at l=0); O2VNet regresses a 3D orientation volume against the ground-truth volume with content, style, projection, and Laplacian losses (Eqs. 3, 6-10); V2VNet updates a rotated volume under additional sketch constraints with the same supervised losses plus an orientation-preserving term (Eqs. 11-12). The training sketch maps are produced by tracing the ground-truth dense orientation map (Sec. 4.3), which is a sparse-to-dense data-generation strategy: the sparse strokes are a subset of the dense field, so the network is asked to infer more information than the input contains. That is a domain-shift/generalization limitation, not circularity, because the dense output is not defined to be the input. Self-citations such as Hair-GAN [9] and example-based synthesis [50] appear as architectural inspiration, data-preprocessing tools, or quantitative baselines; the central claim does not reduce to any of these citations. The pilot study is anecdotal, and the Sec. 4.3 tracing description appears internally inconsistent (choosing the minimum dot-product after thresholding below 0.5), but both are correctness/robustness concerns outside the circularity taxonomy. No fitted parameter is renamed as a prediction, and no derivation is equivalent to its input by construction.

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

The central claim rests on standard supervised learning assumptions plus the domain-specific assumption that synthetic trace-generated sketches match real user drawing behavior. No invented entities are introduced. Free parameters are training hyperparameters and thresholds chosen by hand; they affect performance but are not fitted to validation data.

free parameters (7)
  • content loss weight alpha (S2ONet) = 0.01
    Hand-set weighting for content loss in S2ONet generator loss (Eq. 1).
  • style loss weight beta (S2ONet) = 5
    Hand-set weighting for style loss in S2ONet generator loss (Eq. 1).
  • gradient penalty weight lambda = 10
    Standard WGAN-GP penalty weight used in discriminators (Eq. 4).
  • O2VNet loss weights iota, kappa, gamma, delta, epsilon = 0.01, 5, 0.1, 0.5, 2e-5
    Hand-set weights in O2VNet generator loss (Eq. 10).
  • V2VNet original-field loss weight zeta = 0.1
    Hand-set weight for preserving previous 3D orientation field (Eq. 12).
  • sketch tracing dot-product threshold rho = 0.5
    Threshold in sketch map generation (Sec. 4.3) controlling when neighboring pixels are candidates.
  • hair growing direction threshold theta = 150 degrees
    Threshold in hair growing algorithm (Sec. 5.3) controlling strand continuity.
assumptions (5)
  • domain assumption The synthetic sketch traces generated from ground-truth orientation fields are representative of real user sketches.
    Used throughout training data generation (Sec. 4.3); the network never sees real freehand input during training.
  • domain assumption A dense 2D orientation field is sufficient as an intermediate representation to bridge sparse sketches and 3D orientation volumes.
    Motivates the S2ONet design in Sec. 4; based on the authors' claim that direct sketch-to-3D learning fails, citing [30].
  • domain assumption The 3D orientation field plus the bust depth map contains enough information to reconstruct strand-level hair.
    Used in hair synthesis (Sec. 5.3) and O2VNet; inherits the vector-field representation from [9].
  • domain assumption The 653-model dataset covers the hairstyle space users will request.
    The paper states in Sec. 7.4 that unseen styles like polystrip cannot be produced, acknowledging the dataset's coverage limits.
  • domain assumption WGAN-GP training with the stated loss weights converges to a useful solution.
    Assumed for all three networks; standard practice but not proven for these specific architectures.

how reviews work

0 comments
Cite this review

Pith. "Pith review of DeepSketchHair: Deep Sketch-based 3D Hair Modeling." pith.science (2026). https://pith.science/paper/UV6MC5F5

@misc{pith2026190807198,
  author       = {Pith},
  title        = {Pith review of: DeepSketchHair: Deep Sketch-based 3D Hair Modeling},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/UV6MC5F5}},
  note         = {Machine review of arXiv:1908.07198}
}
read the original abstract

We present sketchhair, a deep learning based tool for interactive modeling of 3D hair from 2D sketches. Given a 3D bust model as reference, our sketching system takes as input a user-drawn sketch (consisting of hair contour and a few strokes indicating the hair growing direction within a hair region), and automatically generates a 3D hair model, which matches the input sketch both globally and locally. The key enablers of our system are two carefully designed neural networks, namely, S2ONet, which converts an input sketch to a dense 2D hair orientation field; and O2VNet, which maps the 2D orientation field to a 3D vector field. Our system also supports hair editing with additional sketches in new views. This is enabled by another deep neural network, V2VNet, which updates the 3D vector field with respect to the new sketches. All the three networks are trained with synthetic data generated from a 3D hairstyle database. We demonstrate the effectiveness and expressiveness of our tool using a variety of hairstyles and also compare our method with prior art.

Figures

Figures reproduced from arXiv: 1908.07198 by the authors.

Figure 1
Figure 1. Our system, DeepSketchHair, takes as input a 2D sketch (a) consisting of a hair contour (red) and a few strokes (blue) indicating the hair growing directions, and generates a realistic 3D hairstyle (b). Users can continue to modify the generated hair model from another view with a new sketch (c), and our system updates the result accordingly (d). high quality or complex geometry. [16], [17] explore vector field base… view at source ↗
Figure 2
Figure 2. The pipeline of our DeepSketchHair framework. The single-view modeling module (top tow) takes as input a user-drawn sketch on top of a bust model (top left), and generates a hair strand model (top right) from a synthesized intermediate 3D orientation field. The multi-view modeling module (bottom row) takes as input the bust model, the currently synthesized hairstyle, and an additional sketch in a rotated view, and g… view at source ↗
Figure 3
Figure 3. Architecture of our S2ONet. For an input pair of sketch image and mask image {IS, IM}, we enclasp them into a 2D tensor of size 128×128×3. For the generator, the contracting part has 3 downsampling convolutional modules with (32, 64, 128) output channels and the expanding part has 3 upsampling decon￾volutional modules with (64, 32, 16) output channels connected with 8 residual blocks. The input of our discriminator … view at source ↗
Figures from the paper (13 more)
Figure 4
Figure 4. Figure 4: Architecture of our O2VNet generator, which takes the input {Λ, D} with the shape of 128 × 128 × 3. The encoder of 2D convolutional part takes three residual blocks with (16, 32, 64) output channels combing with two max pooling layers for down￾sampling, followed by thr…
Figure 5
Figure 5. Figure 5: From left to right: the sketch map generated by rendering [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]
Figure 6
Figure 6. Figure 6: The generator architecture of our V2VNet. The structures of 2D and 3D convolution are similar to that of O2VNet. module. This module takes as input the 3D orientation field Y (after rotation), the bust depth map (after rotation), and the newly specified sketch and hair…
Figure 8
Figure 8. Figure 8: Comparison of our hair growing algorithm to that by Zhang [PITH_FULL_IMAGE:figures/full_fig_p007_8.png]
Figure 9
Figure 9. Figure 9: The user interface. The user can load an existing portrait [PITH_FULL_IMAGE:figures/full_fig_p008_9.png]
Figure 10
Figure 10. Figure 10: Auxiliary tools ((a) cut and (b) local deform) supported [PITH_FULL_IMAGE:figures/full_fig_p008_10.png]
Figure 11
Figure 11. Figure 11: A gallery of modeling results of 3D hairstyles from different inputs. Hairstyles in (a) and (b) are modeled from single images [PITH_FULL_IMAGE:figures/full_fig_p009_11.png]
Figure 12
Figure 12. Figure 12: Representative 3D hairstyles created by 5 novice partici [PITH_FULL_IMAGE:figures/full_fig_p010_12.png]
Figure 14
Figure 14. Figure 14: Comparisons of our method with previous sketch-based [PITH_FULL_IMAGE:figures/full_fig_p010_14.png]
Figure 13
Figure 13. Figure 13: Comparisons of our method with single-view image-based [PITH_FULL_IMAGE:figures/full_fig_p010_13.png]
Figure 15
Figure 15. Figure 15: (Left). It can be seen that without dense 2D feature maps, the details of generated 3D orientation fields are far away from the ground truth, which is consistent with our inference in §4. O2VNet Structure and Loss Function. To present our im￾provement on 3D orientatio…
Figure 16
Figure 16. Figure 16: Ablations study on V2VNet. The method by Delanoy et al. [37] tends to generate vector fields whose values are inconsistent with the ones in the previous view and thus are often chaotic (the middle column). For each method, we show the corresponding results in the prev…
Figure 17
Figure 17. Figure 17: Our method might fail to synthesize hair details due to [PITH_FULL_IMAGE:figures/full_fig_p012_17.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

53 extracted references · 50 canonical work pages

  1. [1]

    A data-driven approach to four-view image-based hair modeling,

    M. Zhang, M. Chai, H. Wu, H. Yang, and K. Zhou, “A data-driven approach to four-view image-based hair modeling,” ACM Trans. Graph., vol. 36, no. 4, pp. 156:1–156:11, 2017

    work page 2017

  2. [2]

    Avatar digitization from a single image for real- time rendering,

    L. Hu, S. Saito, L. Wei, K. Nagano, J. Seo, J. Fursund, I. Sadeghi, C. Sun, Y .-C. Chen, and H. Li, “Avatar digitization from a single image for real- time rendering,” ACM Trans. Graph., vol. 36, no. 6, pp. 195:1–195:14, 2017

    work page 2017

  3. [3]

    Single-view hair modeling for portrait manipulation,

    M. Chai, L. Wang, Y . Weng, Y . Yu, B. Guo, and K. Zhou, “Single-view hair modeling for portrait manipulation,” ACM Trans. Graph. , vol. 31, no. 4, pp. 116:1–116:8, 2012

    work page 2012

  4. [4]

    Dynamic hair manipulation in images and videos,

    M. Chai, L. Wang, Y . Weng, X. Jin, and K. Zhou, “Dynamic hair manipulation in images and videos,” ACM Trans. Graph., vol. 32, no. 4, pp. 75:1–75:8, 2013

    work page 2013

  5. [5]

    Single-view hair modeling using a hairstyle database,

    L. Hu, C. Ma, L. Luo, and H. Li, “Single-view hair modeling using a hairstyle database,” ACM Trans. Graph., vol. 34, no. 4, pp. 125:1–125:9, 2015

    work page 2015

  6. [6]

    Modeling hair from an rgb-d camera,

    M. Zhang, P. Wu, H. Wu, Y . Weng, Y . Zheng, and K. Zhou, “Modeling hair from an rgb-d camera,” ACM Trans. Graph. , vol. 37, no. 6, pp. 205:1–205:10, 2018

    work page 2018

  7. [7]

    Hairnet: Single-view hair reconstruction using convolutional neural networks,

    Y . Zhou, L. Hu, J. Xing, W. Chen, H.-W. Kung, X. Tong, and H. Li, “Hairnet: Single-view hair reconstruction using convolutional neural networks,” in Proceedings of the European Conference on Computer Vision (ECCV), 2018, pp. 235–251

    work page 2018

  8. [8]

    3d hair synthesis using volumetric variational autoencoders,

    S. Saito, L. Hu, C. Ma, H. Ibayashi, L. Luo, and H. Li, “3d hair synthesis using volumetric variational autoencoders,” ACM Trans. Graph., vol. 37, no. 6, pp. 208:1–208:12, 2018

    work page 2018

Show all 53 references
  1. [9]

    Hair-gan: Recovering 3d hair structure from a single image using generative adversarial networks,

    M. Zhang and Y . Zheng, “Hair-gan: Recovering 3d hair structure from a single image using generative adversarial networks,” Visual Informatics 2019, vol. 3, no. 2, p. to appear, 2019

    2019

  2. [10]

    A trigonal prism-based method for hair image generation,

    Y . Watanabe and Y . Suenaga, “A trigonal prism-based method for hair image generation,” IEEE Computer Graphics and applications , vol. 12, no. 1, pp. 47–53, 1992

    1992

  3. [11]

    Interactive multiresolution hair modeling and editing,

    T.-Y . Kim and U. Neumann, “Interactive multiresolution hair modeling and editing,” ACM Trans. Graph., vol. 21, no. 3, pp. 620–629, 2002

    2002

  4. [12]

    A statistical wisp model and pseudophysical approaches for interactive hairstyle generation,

    B. Choe and H.-S. Ko, “A statistical wisp model and pseudophysical approaches for interactive hairstyle generation,” IEEE Transactions on Visualization and Computer Graphics, vol. 11, no. 2, pp. 160–170, 2005

    2005

  5. [13]

    Sketchy hairstyles,

    X. Mao, S. Isobe, K.-i. Anjyo, and A. Imamiya, “Sketchy hairstyles,” in International 2005 Computer Graphics, 2005, pp. 142–147

    2005

  6. [14]

    A sketching interface for modeling and editing hairstyles,

    S. Malik, “A sketching interface for modeling and editing hairstyles,” in SBM, 2005, pp. 185–194

    2005

  7. [15]

    Realistic hair from a sketch,

    J. Wither, F. Bertails, and M.-P. Cani, “Realistic hair from a sketch,” in IEEE International Conference on Shape Modeling and Applications 2007 (SMI’07), 2007, pp. 33–42

    2007

  8. [16]

    Modeling realistic virtual hairstyles,

    Y . Yu, “Modeling realistic virtual hairstyles,” in Proceedings Ninth Pacific Conference on Computer Graphics and Applications. Pacific Graphics 2001, 2001, pp. 295–304

    2001

  9. [17]

    Sketching hairstyles,

    H. Fu, Y . Wei, C.-L. Tai, and L. Quan, “Sketching hairstyles,” in Pro- ceedings of the 4th Eurographics workshop on Sketch-based interfaces and modeling, 2007, pp. 31–36

    2007

  10. [18]

    Generative adversarial nets,

    I. Goodfellow, J. Pouget-Abadie, M. Mirza, B. Xu, D. Warde-Farley, S. Ozair, A. Courville, and Y . Bengio, “Generative adversarial nets,” in Advances in neural information processing systems , 2014, pp. 2672– 2680

    2014

  11. [19]

    A survey on hair modeling: Styling, simulation, and rendering,

    K. Ward, F. Bertails, T.-Y . Kim, S. R. Marschner, M.-P. Cani, and M. C. Lin, “A survey on hair modeling: Styling, simulation, and rendering,” IEEE transactions on visualization and computer graphics, vol. 13, no. 2, pp. 213–234, 2007

    2007

  12. [20]

    A survey of image-based techniques for hair model- ing,

    Y . Bao and Y . Qi, “A survey of image-based techniques for hair model- ing,” IEEE Access, vol. 6, pp. 18 670–18 684, 2018

    2018

  13. [21]

    A system of 3d hair style synthesis based on the wisp model,

    L.-H. Chen, S. Saeyor, H. Dohi, and M. Ishizuka, “A system of 3d hair style synthesis based on the wisp model,” The Visual Computer, vol. 15, no. 4, pp. 159–170, 1999

    1999

  14. [22]

    The cluster hair model,

    X. D. Yang, Z. Xu, J. Yang, and T. Wang, “The cluster hair model,” Graphical Models, vol. 62, no. 2, pp. 85–103, 2000

    2000

  15. [23]

    V-hairstudio: an interactive tool for hair design,

    Z. Xu and X. D. Yang, “V-hairstudio: an interactive tool for hair design,” IEEE Computer Graphics and Applications , vol. 21, no. 3, pp. 36–43, 2001

    2001

  16. [24]

    Generation of 3d hair model from 2d image using image processing,

    K. W. Ming, T. Hiroki, and N. Masayuki, “Generation of 3d hair model from 2d image using image processing,” inApplications of Digital Image Processing XIX, vol. 2847, 1996, pp. 303–312. IEEE TRANSACTIONS ON VISUALIZATION AND COMPUTER GRAPHICS, VOL. XX, NO. XX. 13

    1996

  17. [25]

    Capture of hair geometry from multiple images,

    S. Paris, H. M. Brice ˜no, and F. X. Sillion, “Capture of hair geometry from multiple images,” ACM Trans. Graph., vol. 23, no. 3, pp. 712–719, 2004

    2004

  18. [26]

    Modeling hair from multiple views,

    Y . Wei, E. Ofek, L. Quan, and H.-Y . Shum, “Modeling hair from multiple views,” ACM Trans. Graph., vol. 24, no. 3, pp. 816–820, 2005

    2005

  19. [27]

    High- quality hair modeling from a single portrait photo,

    M. Chai, L. Luo, K. Sunkavalli, N. Carr, S. Hadap, and K. Zhou, “High- quality hair modeling from a single portrait photo,” ACM Trans. Graph., vol. 34, no. 6, pp. 204:1–204:10, 2015

    2015

  20. [28]

    Autohair: Fully automatic hair modeling from a single image,

    M. Chai, T. Shao, H. Wu, Y . Weng, and K. Zhou, “Autohair: Fully automatic hair modeling from a single image,” ACM Trans. Graph. , vol. 35, no. 4, pp. 116:1–116:12, 2016

    2016

  21. [29]

    Auto-encoding variational bayes,

    P. K. Diederik, M. Welling et al., “Auto-encoding variational bayes,” in Proceedings of the International Conference on Learning Representa- tions (ICLR), 2014

    2014

  22. [30]

    Robust flow- guided neural prediction for sketch-based freeform surface modeling,

    C. Li, H. Pan, Y . Liu, X. Tong, A. Sheffer, and W. Wang, “Robust flow- guided neural prediction for sketch-based freeform surface modeling,” ACM Trans. Graph., vol. 37, no. 6, pp. 238:1–238:12, 2018

    2018

  23. [31]

    Teddy: a sketching interface for 3d freeform design,

    T. Igarashi, T. Igarashi, S. Matsuoka, and H. Tanaka, “Teddy: a sketching interface for 3d freeform design,” in Proceedings of the 26th Annual Conference on Computer Graphics and Interactive Techniques , 1999, pp. 409–416

    1999

  24. [32]

    Plushie: An interactive design system for plush toys,

    Y . Mori and T. Igarashi, “Plushie: An interactive design system for plush toys,” ACM Trans. Graph., vol. 26, no. 3, 2007

    2007

  25. [33]

    Fibermesh: Designing freeform surfaces with 3d curves,

    A. Nealen, T. Igarashi, O. Sorkine, and M. Alexa, “Fibermesh: Designing freeform surfaces with 3d curves,” ACM Trans. Graph., vol. 26, no. 3, 2007

    2007

  26. [34]

    3-sweep: Extracting editable objects from a single photo,

    T. Chen, Z. Zhu, A. Shamir, S.-M. Hu, and D. Cohen-Or, “3-sweep: Extracting editable objects from a single photo,” ACM Trans. Graph. , vol. 32, no. 6, pp. 195:1–195:10, 2013

    2013

  27. [35]

    Geose- mantic snapping for sketch-based modeling,

    A. Shtof, A. Agathos, Y . Gingold, A. Shamir, and D. Cohen-Or, “Geose- mantic snapping for sketch-based modeling,” in Computer graphics forum, vol. 32, no. 2pt2, 2013, pp. 245–253

    2013

  28. [36]

    Shape synthesis from sketches via procedural models and convolutional networks,

    H. Huang, E. Kalogerakis, E. Yumer, and R. Mech, “Shape synthesis from sketches via procedural models and convolutional networks,” IEEE transactions on visualization and computer graphics , vol. 23, no. 8, pp. 2003–2013, 2016

    2003

  29. [37]

    3d sketching using multi-view deep volumetric prediction,

    J. Delanoy, M. Aubry, P. Isola, A. A. Efros, and A. Bousseau, “3d sketching using multi-view deep volumetric prediction,” Proc. ACM Comput. Graph. Interact. Tech., vol. 1, no. 1, pp. 21:1–21:22, 2018

    2018

  30. [38]

    3d shape reconstruction from sketches via multi-view convolutional networks,

    Z. Lun, M. Gadelha, E. Kalogerakis, S. Maji, and R. Wang, “3d shape reconstruction from sketches via multi-view convolutional networks,” in 2017 International Conference on 3D Vision (3DV), 2017

    2017

  31. [39]

    Interactive sketch-based normal map generation with deep neural networks,

    W. Su, D. Du, X. Yang, S. Zhou, and H. Fu, “Interactive sketch-based normal map generation with deep neural networks,” Proc. ACM Comput. Graph. Interact. Tech., vol. 1, no. 1, pp. 22:1–22:17, 2018

    2018

  32. [40]

    Deepsketch2face: A deep learning based sketching system for 3d face and caricature modeling,

    X. Han, C. Gao, and Y . Yu, “Deepsketch2face: A deep learning based sketching system for 3d face and caricature modeling,” ACM Trans. Graph., vol. 36, no. 4, pp. 126:1–126:12, 2017

    2017

  33. [41]

    Image-to-image translation with conditional adversarial networks,

    P. Isola, J.-Y . Zhu, T. Zhou, and A. A. Efros, “Image-to-image translation with conditional adversarial networks,” CVPR, 2017

    2017

  34. [42]

    Improved training of wasserstein gans,

    I. Gulrajani, F. Ahmed, M. Arjovsky, V . Dumoulin, and A. C. Courville, “Improved training of wasserstein gans,” inAdvances in Neural Informa- tion Processing Systems, 2017, pp. 5767–5777

    2017

  35. [43]

    Scribbler: Controlling deep image synthesis with sketch and color,

    P. Sangkloy, J. Lu, C. Fang, F. Yu, and J. Hays, “Scribbler: Controlling deep image synthesis with sketch and color,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition , 2017, pp. 5400–5409

    2017

  36. [44]

    Two-stage sketch colorization,

    L. Zhang, C. Li, T.-T. Wong, Y . Ji, and C. Liu, “Two-stage sketch colorization,” ACM Trans. Graph. , vol. 37, no. 6, pp. 261:1–261:14, 2018

    2018

  37. [45]

    Image style transfer using convolutional neural networks,

    L. A. Gatys, A. S. Ecker, and M. Bethge, “Image style transfer using convolutional neural networks,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2016, pp. 2414–2423

    2016

  38. [46]

    Learning a probabilistic latent space of object shapes via 3d generative-adversarial modeling,

    J. Wu, C. Zhang, T. Xue, W. T. Freeman, and J. B. Tenenbaum, “Learning a probabilistic latent space of object shapes via 3d generative-adversarial modeling,” inAdvances in Neural Information Processing Systems, 2016, pp. 82–90

    2016

  39. [47]

    U-net: Convolutional networks for biomedical image segmentation,

    O. Ronneberger, P. Fischer, and T. Brox, “U-net: Convolutional networks for biomedical image segmentation,” in International Conference on Medical image computing and computer-assisted intervention, 2015, pp. 234–241

    2015

  40. [48]

    Multi-view supervision for single-view reconstruction via differentiable ray consistency,

    S. Tulsiani, T. Zhou, A. A. Efros, and J. Malik, “Multi-view supervision for single-view reconstruction via differentiable ray consistency,” CoRR, vol. abs/1704.06254, 2017

    2017 arXiv

  41. [49]

    Photographic image synthesis with cascaded refinement networks,

    Q. Chen and V . Koltun, “Photographic image synthesis with cascaded refinement networks,” in Proceedings of the IEEE International Confer- ence on Computer Vision, 2017, pp. 1511–1520

    2017

  42. [50]

    Example-based hair geometry synthesis,

    L. Wang, Y . Yu, K. Zhou, and B. Guo, “Example-based hair geometry synthesis,” ACM Trans. Graph., vol. 28, no. 3, pp. 56:1–56:9, 2009

    2009

  43. [51]

    V2v-posenet: V oxel-to-voxel prediction network for accurate 3d hand and human pose estimation from a single depth map,

    G. Moon, J. Yong Chang, and K. Mu Lee, “V2v-posenet: V oxel-to-voxel prediction network for accurate 3d hand and human pose estimation from a single depth map,” inProceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2018, pp. 5079–5088

    2018

  44. [52]

    Displaced dynamic expression regression for real-time facial tracking and animation,

    C. Cao, Q. Hou, and K. Zhou, “Displaced dynamic expression regression for real-time facial tracking and animation,”ACM Trans. Graph., vol. 33, no. 4, pp. 43:1–43:10, 2014

    2014

  45. [53]

    Laplacian surface editing,

    O. Sorkine, D. Cohen-Or, Y . Lipman, M. Alexa, C. R ¨ossl, and H.-P. Seidel, “Laplacian surface editing,” in Proceedings of the 2004 Euro- graphics/ACM SIGGRAPH Symposium on Geometry Processing , ser. SGP ’04, 2004, pp. 175–184. Y uefan Shen is a Ph.D. candidate at the State K...

    2004

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.