SliderGAN: Synthesizing Expressive Face Images by Sliding 3D Blendshape Parameters

Evangelos Ververas; Stefanos Zafeiriou

REVIEW 4 major objections 5 minor 35 references

SliderGAN: Synthesizing Expressive Face Images by Sliding 3D Blendshape Parameters

T0 review · 4 major / 5 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 SliderGAN is a generative model that edits a face photo by sliding continuous 3D blendshape parameters for expression and speech.

desk verdict A genuinely new conditioning signal for face editing, but the evaluation leans heavily on the same 3DMM fitting used to create the pseudo-labels; still worth refereeing. read the letter →

arxiv 1908.09638 v1 pith:RLGN55XN submitted 2019-08-26 cs.CV

Evangelos Ververas , Stefanos Zafeiriou This is my paper · ORCID

classification cs.CV

keywords image-to-imagetranslationgenerativeadversarialnetworkblendshapemodelfacialexpressionsynthesisspeechanimation3Dmorphablefittingrelativisticdiscriminatorfaceediting

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 introduces SliderGAN, a generative adversarial network that edits a face image by conditioning on continuous parameter values of a 3D blendshape model of facial motion, so that moving a slider along one blendshape axis produces smooth expression changes in the image. The authors claim this is the first image-to-image translation approach driven directly by statistical blendshape parameters rather than discrete emotion labels or action-unit intensity estimates, and that it needs no manual annotation because the parameters come from 3D Morphable Model fitting and audio alignment. If correct, SliderGAN provides a single flexible control signal for expression editing, transfer, interpolation, neutralization, and speech animation from a single image, and it outperforms AU-based GANimation on image distance and expression recognition metrics.

What carries the argument

The central object is the linear 3D blendshape model $S(\mathbf{p}) = \bar{s} + U\mathbf{p}$, whose sparse localized deformation components were learned from 4DFAB scans for expression and from LRW-3D audio alignment for speech. SliderGAN uses pseudo-annotations from 3DMM fitting of this model to label 250,000 Emotionet images, and renders 200,000 synthetic image pairs with known parameter values to supervise the generator. The mechanism that carries the argument is the continuous, interpretable parameter vector itself: it is the conditioning signal that drives the attention-based generator, the regression target for the relativistic discriminator, and the interpolation basis for expression transfer.

What would settle it

Render a set of faces from 4D scans with known blendshape coefficients, slide one coefficient of SliderGAN from -1 to 1 on those input images, then refit the output meshes with an independent 3D face reconstruction system that was not used to generate the pseudo-annotations; if the recovered coefficients do not move monotonically with the slider or if identity parameters shift substantially, the claimed continuous slider-to-expression mapping is not established.

Watch

Extended reading notes

Core claim

SliderGAN establishes that a GAN generator, conditioned on a normalized vector of 3D blendshape parameters, can translate an input face image into an output that exhibits the facial motion instantiated by that parameter vector in the blendshape model $S_{\text{exp}}(\mathbf{p}) = \bar{s} + U_{\text{exp}}\mathbf{p}$. The generator outputs a deformation mask and a deformation image that are combined with the original image, and it is trained with adversarial, expression-regression, reconstruction, image-pair, identity, and attention-sparsity losses. A relativistic average discriminator both judges realism and regresses the expression parameters, which doubles as a by-product network for recovering blendshape parameters from images. The paper shows this works for 30 expression components from 4DFAB and 10 speech components from LRW-3D, enabling single- and multi-parameter editing, expression transfer and interpolation, neutralization, and speech animation, and reports that SliderGAN trained with the relativistic discriminator yields lower Image Euclidean Distance than GANimation on 4DFAB and LRW benchmarks.

Load-bearing premise

The load-bearing premise is that the 3DMM fitting that produces the pseudo-annotations, and the audio alignment that produces speech parameters, recover the expression separably from identity, texture, pose, and camera settings, so that the training labels actually encode the facial motion the sliders claim to control.

Editorial extensions

If this is right

Expression editing becomes continuous: sliding any single blendshape parameter between -1 and 1 yields smooth changes in the image, with the zero vector producing a neutral expression.
Expression transfer and interpolation between two images can be done without 3DMM fitting at test time, because the discriminator regresses the blendshape parameters of any target image.
The same architecture handles speech: conditioning on LRW-3D speech blendshape parameters animates lip and mouth motion from a single image, with lower IED than AU-based GANimation.
The by-product regression network recovers blendshape parameters from images with higher consistency than OpenFace AU detection, making blendshape coding a more reliable control signal than action units for this task.
Relativistic GAN training improves texture detail over WGAN-GP in this image-to-image translation setting.

Reading between the lines

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

Because the control signal is a physical face model rather than categorical labels, the same architecture could in principle be steered by any 3DMM parameterization (age, weight, identity) or by other dense physical models, such as body pose or hand articulation, replacing discrete domain labels with continuous physics-driven sliders.
The reliance on a single fitting method for pseudo-annotations suggests a direct test: re-annotate the same training images with an independent 3D face fitter or with manual 4D scan alignment; if SliderGAN's image quality and expression accuracy degrade, the method's success is partly an artifact of the annotation procedure rather than a general property of blendshape conditioning.
The reported identity and speech results imply a practical recipe for talking-head animation from a single image; combining SliderGAN with audio-to-blendshape predictors would produce video avatars whose speech motion is controlled by a compact, interpretable parameter stream rather than by per-frame image regression.

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.

Desk Editor's Note

read the letter

Colleague,

The gist: SliderGAN conditions a GAN on continuous 3D blendshape parameters rather than discrete labels or AUs, and that is a genuinely new control signal for face editing. It also extends to speech animation from a single image. The architecture is mostly assembled from known parts—CycleGAN generator, attention masks, RaGAN, ArcFace loss—but the problem setup is the contribution.

The paper does several things well. The experiments are broad: single and multi-parameter editing, expression transfer and interpolation, discrete expression synthesis, neutralization, and speech synthesis. They built a rendered 4DFAB dataset for quantitative IED evaluation and augmented training with 200k synthetic pairs that have exactly known parameters. The ablation study is proper. The writing is clear, and the authors admit the fitting limitation in Section 4.1.

The soft spots are real but not fatal. The main one is the one the stress-test highlights: the pseudo-labels come from 3DMM fitting, and the quantitative evaluation on 4DFAB uses the same blendshape model for ground truth. If the fitting systematically confounds expression with identity/camera, that bias is invisible in the IED numbers. The synthetic data help but don't eliminate the confound, since the identity mesh used for rendering is itself fitted to the original image. So the 'accuracy' of the slider-to-expression mapping is not independently validated against physical motion. The discrete expression recognition with an off-the-shelf network is an external check, and it supports the claim, but the headline IED comparisons lack error bars and significance tests. Also, the GANimation baseline uses OpenFace AU detections, which the authors themselves call low-accuracy; that comparison may be too kind to SliderGAN. Finally, no code is released, so the reader can't verify the visuals or reproduce the numbers.

All that said, the idea is worth taking seriously. The continuous conditioning on a physical model, with the speech extension, is something previous work didn't do. The paper deserves peer review. Reviewers should ask for independent evaluation of the parameterization—e.g., against 3D scans with known expression, or human perceptual studies—and for error bars on the main comparisons.

Who it's for: anyone working on face editing, expression synthesis, or GAN-based image translation. I'd probably send it out rather than desk reject.

Best.

Referee Report

4 major / 5 minor

Summary. The paper proposes SliderGAN, an image-to-image translation method that edits a face image according to continuous values of a 3D statistical blendshape model of facial expression and speech. The generator is conditioned on a vector of blendshape parameters, and a relativistic discriminator with an auxiliary regression head provides a parameter-consistency signal. Training is semi-supervised: it uses unpaired images whose blendshape parameters are obtained by 3DMM fitting, plus synthetic image pairs rendered from fitted identity/camera models with known random expression values. The authors report qualitative results for single- and multi-parameter editing, expression transfer/interpolation, discrete expression synthesis, neutralization, and combined expression-speech animation, along with quantitative comparisons to GANimation using Image Euclidean Distance on a rendered 4DFAB dataset and on LRW frames, expression recognition accuracy, and parameter-regression consistency.

Significance. If the quantitative claims hold, SliderGAN is a meaningful advance over AU- and discrete-expression-driven face editing: it provides a continuous, physically interpretable control space shared with 3D face models, extends naturally to speech-driven animation, and avoids manual annotation. The introduction of a rendered 4DFAB benchmark with ground-truth blendshape parameters is a useful resource for the community. The paper also demonstrates a concrete benefit of the relativistic discriminator in this task, and the ablation study gives some evidence about which losses matter. However, the load-bearing quantitative evidence currently has gaps in statistical rigor and independence, which the major comments detail.

major comments (4)

[Eq. (15), Tables 1 and 3] The Image Euclidean Distance formula as printed is identically zero: the second norm inside the double sum is written as (||xj−xj||^2), which vanishes for every j, so the reported IED values in Tables 1 and 3 cannot be reproduced from the given definition. This appears to be a typo for (||xj−yj||^2), but as published the metric is undefined and the central quantitative comparison is not reproducible.
[Tables 1 and 3] The IED comparisons report only point estimates computed over 10,000 pairs, with no standard deviations, confidence intervals, or significance tests. The differences between GANimation (1.04e−02), SliderGAN-WGP (7.93e−03), and SliderGAN-RaD (6.84e−03) in Table 1, and similarly in Table 3, may or may not be meaningful given the variance of the metric. The claim that SliderGAN outperforms GANimation therefore needs error bars or a statistical test.
[Section 4.1] The manuscript acknowledges that 3DMM fitting 'can be mistakenly explained by the identity, texture or camera model of the 3DMM,' yet the quantitative evaluation on the rendered 4DFAB dataset derives ground-truth parameters by the same projection p = U^T(S−mean) from the same blendshape model. A systematic fitting bias is therefore invisible in Table 1, and the synthetic-pair mitigation in Section 4.1 reuses the same identity/camera fits for the input images, so it does not independently validate the slider-to-expression mapping.
[Section 4.6, Table 4] The second row of Table 4 measures consistency between the target parameters ptrg and D(G(Iorg, ptrg)) on images generated by SliderGAN. Because the generator is explicitly trained with Lexp,G (Eq. 8) to make this regression match ptrg, this is a measure of how well the model satisfies its own training objective, not an independent estimate of expression accuracy. It should be validated against 3DMM fitting of the generated images or against manual/landmark-based measures before claiming that D 'extracts expression parameters with high accuracy.'

minor comments (5)

[Eq. (7)] There is an extra closing parenthesis in Lexp,D = 1/N ||D(Iorg) − porg)||^2; the formula should read ||D(Iorg) − porg||^2.
[Section 4.7] The ablation text says 'c) LG does not include bothLid andLid' and later 'both Lid andLid are omitted'; the intended meaning is clearly 'both Lid and Lgen,' and the figure caption 'without Lid+Lgen' should be harmonized with the text.
[Abstract] The phrase 'and particular using the learning paradigm' should read 'and in particular using.'
[Section 4.1] The description of the synthetic data states that 200,000 images are rendered from 10,000 identities with 20 expressions each; providing details on how the rendered images are aligned and whether any post-processing is applied would help reproducibility.
[Section 3.1] The paper calls the method 'annotation-free,' but it uses pseudo-annotations from 3DMM fitting; the distinction between manual annotation and automatic pseudo-labels should be made explicit in the introduction to avoid overstatement.

Circularity Check

2 steps flagged · score 6.0 of 10

The discriminator-consistency evaluation in Table 4 restates the training objective Lexp,G; the central editing claim retains independent IED and recognition benchmarks.

fitted input called prediction [Section 4.6, Eq. (8), Table 4 row 2]
"Lexp,G = 1/N||D(G(Iorg, ptrg))-ptrg||^2. ... b) we test the consistency of our model and calculate the error between some target parameters ptrg and those regressed by D on a manipulated image which was translated to expression ptrg by SliderGAN-RaD."

This row-2 error is exactly the quantity minimized by Eq. (8) during training: D is trained to output ptrg when shown G(Iorg, ptrg). Reporting a low value of this residual as a consistency result is therefore measuring the training objective itself, not an independent check that the synthesized image carries the expression ptrg. The only external validity of ptrg comes from the 3DMM fitting or rendering pipeline; the Table 4 row-2 number adds no evidence beyond the fit.
fitted input called prediction [Section 4.6, Eq. (7), Table 4 row 1]
"Lexp,D = 1/N||D(Iorg)-porg)||^2. ... a) we calculate the error between parameters recovered by 3DMM fitting and those regressed by D on the same image (Table 4 row 1)."

The regression head of D is trained under Eq. (7) to match porg produced by the same 3DMM fitting pipeline that later supplies the reference values in Table 4 row 1. A low row-1 error on held-out images shows that D generalizes the fitting function to new images, not that the fitted parameters are physically accurate or that SliderGAN generates correct expressions. The benchmark is self-referential with respect to the pseudo-label source.

full rationale

SliderGAN's central editing claim is not fully circular: the IED experiments on rendered 4DFAB images and the discrete-expression recognition results use external benchmarks, and the synthetic-pair training with Lgen is a genuine supervised signal from rendered images. However, the '3D Expression Reconstruction' evaluation in Section 4.6 is partly self-referential. Table 4 row 2 measures ||ptrg - D(G(Iorg, ptrg))||, which is exactly the Lexp,G term minimized in Eq. (8), so the reported consistency is a restatement of the training fit rather than independent evidence. Table 4 row 1 similarly compares D to the same 3DMM fitting procedure that generated D's pseudo-labels via Eq. (7). The paper itself admits in Section 4.1 that 'parameter values are not always consistent as small variations in expression can be mistakenly explained by the identity, texture or camera model of the 3DMM,' which means these internal agreements cannot validate physical expression accuracy. The self-citations to [9], [10], [11], and [30] are data and tools with external provenance, not an unverified uniqueness argument, so they do not by themselves raise the score. Because external benchmarks exist, the circularity is partial rather than total.

Assumptions & free parameters 4 free parameters · 6 assumptions · 0 invented entities

The central claim rests on a pipeline of pre-existing statistical face models and fitting procedures. The model introduces no new physical entities, but it depends heavily on the accuracy of 3DMM pseudo-labels, the coverage of the 4DFAB and LRW-3D blendshape models, and the validity of ArcFace as an identity measure. The hand-chosen loss weights and component counts are additional free parameters that are not derived from first principles.

free parameters (4)

Loss weights lambda_adv, lambda_exp, lambda_rec, lambda_gen, lambda_id, lambda_att = 30, 1000, 10, 10, 4, 0.3
Chosen by hand in Section 3.2; they directly determine the balance between realism, expression accuracy, reconstruction, identity, and attention sparsity.
Number of blendshape components = 30 expression (4DFAB), 10 speech (LRW-3D)
Selected as the most significant components; affects the expressiveness of the conditioning vector and the capacity of the discriminator regression.
Image resolution = 128x128
All training and synthesis are performed at 128x128, which caps detail and is a design choice rather than a derived quantity.
Two-stage training schedule = 20 epochs paired + 40 epochs unpaired
The schedule is chosen empirically and is a free procedural parameter that affects final image quality.

assumptions (6)

domain assumption 3DMM fitting of Booth et al. [9] recovers expression parameters from images accurately enough to serve as pseudo-labels for training.
Section 2.2 and Section 4.1: the entire annotation pipeline uses fitted parameters pexp as ground truth; noise or bias in fitting propagates into the generator.
domain assumption The 4DFAB blendshape model spans the facial motion space relevant for expression editing and rendering.
Section 2.1 and Section 4.1: the model is used both to define the target parameter space and to render pseudo-ground-truth images; if it misses motion modes, SliderGAN cannot synthesize them.
domain assumption LRW-3D speech blendshape parameters accurately encode lip and jaw motion during speech.
Section 4.5: speech synthesis is conditioned on these parameters, and LRW-3D is the source of ground-truth speech deformations.
domain assumption ArcFace embeddings are a valid proxy for identity preservation.
Eq.(11) minimizes cosine distance between embeddings; this assumes identity is well captured by those embeddings.
domain assumption The linear blendshape model Sexp(p) = mean + U_exp p adequately represents facial deformations in images.
Eq.(3) and Section 3.1 define the problem on a linear model; non-linear deformations and expression-identity coupling are not modeled.
domain assumption Rendered 3D faces from 4DFAB are realistic enough to serve as ground truth for IED and as synthetic training pairs.
Section 4.1: 300,000 rendered images and 200,000 synthetic warped images are used for training and quantitative evaluation; rendering quality bounds the fidelity of both.

how reviews work

0 comments

Cite this review

Pith. "Pith review of SliderGAN: Synthesizing Expressive Face Images by Sliding 3D Blendshape Parameters." pith.science (2026). https://pith.science/paper/RLGN55XN

@misc{pith2026190809638,
  author       = {Pith},
  title        = {Pith review of: SliderGAN: Synthesizing Expressive Face Images by Sliding 3D Blendshape Parameters},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/RLGN55XN}},
  note         = {Machine review of arXiv:1908.09638}
}

read the original abstract

Image-to-image (i2i) translation is the dense regression problem of learning how to transform an input image into an output using aligned image pairs. Remarkable progress has been made in i2i translation with the advent of Deep Convolutional Neural Networks (DCNNs) and particular using the learning paradigm of Generative Adversarial Networks (GANs). In the absence of paired images, i2i translation is tackled with one or multiple domain transformations (i.e., CycleGAN, StarGAN etc.). In this paper, we study a new problem, that of image-to-image translation, under a set of continuous parameters that correspond to a model describing a physical process. In particular, we propose the SliderGAN which transforms an input face image into a new one according to the continuous values of a statistical blendshape model of facial motion. We show that it is possible to edit a facial image according to expression and speech blendshapes, using sliders that control the continuous values of the blendshape model. This provides much more flexibility in various tasks, including but not limited to face editing, expression transfer and face neutralisation, comparing to models based on discrete expressions or action units.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

35 extracted references · 26 canonical work pages

[1]

3693–3703 (2018)

Alami Mejjati, Y., Richardt, C., Tompkin, J., Cosker, D., Kim, K.I.: Unsupervised attention- guided image-to-image translation pp. 3693–3703 (2018). URL http://papers.nips.cc/paper/ 7627-unsupervised-attention-guided-image-to-image-translation. pdf

work page 2018
[2]

Amos, B., Ludwiczuk, B., Satyanarayanan, M.: Openface: A general-purpose face recognition library with mobile applications. Tech. rep., CMU-CS-16-118, CMU School of Computer Science (2016)

work page 2016
[3]

In: Proceedings of the 34th International Conference on Machine Learning, ICML 2017, Sydney, NSW, Australia, 6-11 August 2017, pp

Arjovsky, M., Chintala, S., Bottou, L.: Wasserstein gen- erative adversarial networks. In: Proceedings of the 34th International Conference on Machine Learning, ICML 2017, Sydney, NSW, Australia, 6-11 August 2017, pp. 214–223 (2017). URL http://proceedings.mlr.press/ v70/arjovsky17a.html

work page 2017
[4]

Foundations and Trends in Machine Learning 4(1), 1–106 (2012)

Bach, F., Jenatton, R., Mairal, J., Obozinski, G.: Opti- mization with sparsity-inducing penalties. Foundations and Trends in Machine Learning 4(1), 1–106 (2012). DOI 10.1561/2200000015. URL http://dx.doi.org/10. 1561/2200000015

doi:10.1561/2200000015 2012
[5]

In: 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp

Benitez-Quiroz, C.F., Srinivasan, R., Martinez, A.M.: Emotionet: An accurate, real-time algorithm for the au- tomatic annotation of a million facial expressions in the wild. In: 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 5562–5570 (2016). DOI 10.1109/CVPR.2016.600

doi:10.1109/cvpr.2016.600 2016
[6]

In: ICCV, pp

Benitez-Quiroz, C.F., Wang, Y., Martinez, A.M.: Recog- nition of action units in the wild with deep nets and a new global-local loss. In: ICCV, pp. 3990–3999 (2017)

work page 2017
[7]

IEEE transactions on pattern analysis and machine in- telligence (2018)

Benitez-Quiroz, F., Srinivasan, R., Martinez, A.M.: Dis- criminant functional learning of color features for the recognition of facial action units and their intensities. IEEE transactions on pattern analysis and machine in- telligence (2018)

work page 2018
[8]

In: Proceedings of the IEEE Confer- ence on ComputerVision and Pattern Recognition (2017)

Booth, J., Antonakos, E., Ploumpis, S., Trigeorgis, G., Panagakis, Y., Zafeiriou, S., et al.: 3d face morphable models in-the-wild. In: Proceedings of the IEEE Confer- ence on ComputerVision and Pattern Recognition (2017)

work page 2017

Show all 35 references

[9]

IEEE Transactions on Pattern Analysis and Machine In- telligence (2018)

Booth, J., Roussos, A., Ververas, E., Antonakos, E., Poumpis, S., Panagakis, Y., Zafeiriou, S.P.: 3d recon- struction of” in-the-wild” faces in images and videos. IEEE Transactions on Pattern Analysis and Machine In- telligence (2018)

2018
[10]

In: 2016 IEEE Conference on Computer Vision and Pat- tern Recognition (CVPR), pp

Booth, J., Roussos, A., Zafeiriou, S., Ponniahy, A., Dun- away, D.: A 3d morphable model learnt from 10,000 faces. In: 2016 IEEE Conference on Computer Vision and Pat- tern Recognition (CVPR), pp. 5543–5552 (2016). DOI 10.1109/CVPR.2016.598

2016 doi
[11]

In: 2018 IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2018)

Cheng, S., Kotsia, I., Pantic, M., Zafeiriou, S.: 4dfab: A large scale 4d database for facial expression analysis and biometric applications. In: 2018 IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2018). Salt Lake City, Utah, US (2018)

2018
[12]

In: The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2018)

Choi, Y., Choi, M., Kim, M., Ha, J.W., Kim, S., Choo, J.: Stargan: Uniﬁed generative adversarial networks for multi-domain image-to-image translation. In: The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2018)

2018
[13]

In: Asian Conference on Computer Vision (2016)

Chung, J.S., Zisserman, A.: Lip reading in the wild. In: Asian Conference on Computer Vision (2016)

2016
[14]

arXiv:1801.07698 (2018)

Deng, J., Guo, J., Zafeiriou, S.: Arcface: Additive angular margin loss for deep face recognition. arXiv:1801.07698 (2018)

2018 arXiv
[15]

A human face (2002)

Ekman, P.: Facial action coding system (facs). A human face (2002)

2002
[16]

Goodfellow, I., Pouget-Abadie, J., Mirza, M., Xu, B., Warde-Farley, D., Ozair, S., Courville, A., Ben- gio, Y.: Generative adversarial nets. In: Z. Ghahra- mani, M. Welling, C. Cortes, N.D. Lawrence, K.Q. Weinberger (eds.) Advances in Neural Information Processing Systems 27, ...

2014
[17]

Gulrajani, I., Ahmed, F., Arjovsky, M., Dumoulin, V., Courville, A.C.: Improved training of wasser- stein gans. In: I. Guyon, U.V. Luxburg, S. Ben- gio, H. Wallach, R. Fergus, S. Vishwanathan, R. Gar- nett (eds.) Advances in Neural Information Process- ing Systems 30, pp. 5767...

2017
[18]

CVPR (2017)

Isola, P., Zhu, J.Y., Zhou, T., Efros, A.A.: Image-to- image translation with conditional adversarial networks. CVPR (2017)

2017
[19]

In: Interna- tional Conference on Learning Representations (2019)

Jolicoeur-Martineau, A.: The relativistic discriminator: a key element missing from standard GAN. In: Interna- tional Conference on Learning Representations (2019). URL https://openreview.net/forum?id=S1erHoR5t7

2019
[20]

CoRR abs/1412.6980 (2014)

Kingma, D.P., Ba, J.: Adam: A method for stochastic optimization. CoRR abs/1412.6980 (2014). URL http: //arxiv.org/abs/1412.6980

2014 arXiv
[21]

CoRR abs/1610.05586 (2016)

Li, M., Zuo, W., Zhang, D.: Deep identity-aware trans- fer of facial attributes. CoRR abs/1610.05586 (2016). URL http://arxiv.org/abs/1610.05586

2016 arXiv
[22]

In: Computer Vision and Pattern Recognition (CVPR), 2017 IEEE Conference on, pp

Li, S., Deng, W., Du, J.: Reliable crowdsourcing and deep locality-preserving learning for expression recognition in 18 Evangelos Ververas ⋆, Stefanos Zafeiriou † the wild. In: Computer Vision and Pattern Recognition (CVPR), 2017 IEEE Conference on, pp. 2584–2593. IEEE (2017)

2017
[23]

ACM Transactions on Graphics (TOG) 32(6), 179 (2013)

Neumann, T., Varanasi, K., Wenger, S., Wacker, M., Magnor, M., Theobalt, C.: Sparse localized deformation components. ACM Transactions on Graphics (TOG) 32(6), 179 (2013)

2013
[24]

CoRR abs/1611.06355 (2016)

Perarnau, G., van de Weijer, J., Raducanu, B., ´Alvarez, J.M.: Invertible conditional gans for image editing. CoRR abs/1611.06355 (2016). URL http://arxiv.org/abs/ 1611.06355

2016 arXiv
[25]

In: Proceedings of the European Conference on Computer Vision (ECCV) (2018)

Pumarola, A., Agudo, A., Martinez, A., Sanfeliu, A., Moreno-Noguer, F.: Ganimation: Anatomically-aware fa- cial animation from a single image. In: Proceedings of the European Conference on Computer Vision (ECCV) (2018)

2018
[26]

In: Computer Vision and Pattern Recognition (CVPR), 2017 IEEE Conference on, pp

Richardson, E., Sela, M., Or-El, R., Kimmel, R.: Learn- ing detailed face reconstruction from a single image. In: Computer Vision and Pattern Recognition (CVPR), 2017 IEEE Conference on, pp. 5553–5562. IEEE (2017)

2017
[27]

ACM Transactions on Graphics (TOG) 36(4), 95 (2017)

Suwajanakorn, S., Seitz, S.M., Kemelmacher-Shlizerman, I.: Synthesizing obama: learning lip sync from audio. ACM Transactions on Graphics (TOG) 36(4), 95 (2017)

2017
[28]

arXiv preprint arXiv:1712.02859 2 (2017)

Tewari, A., Zollh¨ ofer, M., Garrido, P., Bernard, F., Kim, H., P´ erez, P., Theobalt, C.: Self-supervised multi-level face model learning for monocular reconstruction at over 250 hz. arXiv preprint arXiv:1712.02859 2 (2017)

2017 arXiv
[29]

arXiv preprint arXiv:1804.03786 (2018)

Tran, L., Liu, X.: Nonlinear 3d face morphable model. arXiv preprint arXiv:1804.03786 (2018)

2018 arXiv
[30]

Tzirakis, P., Papaioannou, A., Lattas, A., Tarasiou, M., Schuller, B., Zafeiriou, S.: Synthesising 3d facial motion from”in-the-wild”speech (2019)

2019
[31]

IEEE Trans

Wang, L., Zhang, Y., Feng, J.: On the euclidean distance of images. IEEE Trans. Pattern Anal. Mach. Intell.27(8), 1334–1339 (2005). DOI 10.1109/TPAMI.2005.165. URL https://doi.org/10.1109/TPAMI.2005.165

2005 doi
[32]

In: The European Conference on Computer Vision Workshops (ECCVW) (2018)

Wang, X., Yu, K., Wu, S., Gu, J., Liu, Y., Dong, C., Qiao, Y., Loy, C.C.: Esrgan: Enhanced super-resolution gener- ative adversarial networks. In: The European Conference on Computer Vision Workshops (ECCVW) (2018)

2018
[33]

In: Proc

Wiles, O., Koepke, A.S., Zisserman, A.: X2face: A net- work for controlling face generation using images, audio, and pose codes. In: Proc. ECCV (2018)

2018
[34]

IEEE Trans- actions on Signal Processing 57(7), 2479–2493 (2009)

Wright, S.J., Nowak, R.D., Figueiredo, M.A.T.: Sparse reconstruction by separable approximation. IEEE Trans- actions on Signal Processing 57(7), 2479–2493 (2009). DOI 10.1109/TSP.2009.2016892

2009
[35]

In: Computer Vision (ICCV), 2017 IEEE International Conference on (2017)

Zhu, J.Y., Park, T., Isola, P., Efros, A.A.: Unpaired image-to-image translation using cycle-consistent adver- sarial networkss. In: Computer Vision (ICCV), 2017 IEEE International Conference on (2017)

2017

Pith tools

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

[1] [1]

3693–3703 (2018)

Alami Mejjati, Y., Richardt, C., Tompkin, J., Cosker, D., Kim, K.I.: Unsupervised attention- guided image-to-image translation pp. 3693–3703 (2018). URL http://papers.nips.cc/paper/ 7627-unsupervised-attention-guided-image-to-image-translation. pdf

work page 2018

[2] [2]

Amos, B., Ludwiczuk, B., Satyanarayanan, M.: Openface: A general-purpose face recognition library with mobile applications. Tech. rep., CMU-CS-16-118, CMU School of Computer Science (2016)

work page 2016

[3] [3]

In: Proceedings of the 34th International Conference on Machine Learning, ICML 2017, Sydney, NSW, Australia, 6-11 August 2017, pp

Arjovsky, M., Chintala, S., Bottou, L.: Wasserstein gen- erative adversarial networks. In: Proceedings of the 34th International Conference on Machine Learning, ICML 2017, Sydney, NSW, Australia, 6-11 August 2017, pp. 214–223 (2017). URL http://proceedings.mlr.press/ v70/arjovsky17a.html

work page 2017

[4] [4]

Foundations and Trends in Machine Learning 4(1), 1–106 (2012)

Bach, F., Jenatton, R., Mairal, J., Obozinski, G.: Opti- mization with sparsity-inducing penalties. Foundations and Trends in Machine Learning 4(1), 1–106 (2012). DOI 10.1561/2200000015. URL http://dx.doi.org/10. 1561/2200000015

doi:10.1561/2200000015 2012

[5] [5]

In: 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp

Benitez-Quiroz, C.F., Srinivasan, R., Martinez, A.M.: Emotionet: An accurate, real-time algorithm for the au- tomatic annotation of a million facial expressions in the wild. In: 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 5562–5570 (2016). DOI 10.1109/CVPR.2016.600

doi:10.1109/cvpr.2016.600 2016

[6] [6]

In: ICCV, pp

Benitez-Quiroz, C.F., Wang, Y., Martinez, A.M.: Recog- nition of action units in the wild with deep nets and a new global-local loss. In: ICCV, pp. 3990–3999 (2017)

work page 2017

[7] [7]

IEEE transactions on pattern analysis and machine in- telligence (2018)

Benitez-Quiroz, F., Srinivasan, R., Martinez, A.M.: Dis- criminant functional learning of color features for the recognition of facial action units and their intensities. IEEE transactions on pattern analysis and machine in- telligence (2018)

work page 2018

[8] [8]

In: Proceedings of the IEEE Confer- ence on ComputerVision and Pattern Recognition (2017)

Booth, J., Antonakos, E., Ploumpis, S., Trigeorgis, G., Panagakis, Y., Zafeiriou, S., et al.: 3d face morphable models in-the-wild. In: Proceedings of the IEEE Confer- ence on ComputerVision and Pattern Recognition (2017)

work page 2017