Filters








4,207 Hits in 6.9 sec

Conditional Adversarial Synthesis of 3D Facial Action Units

Zhilei Liu, Guoxian Song, Jianfei Cai, Tat-Jen Cham, Juyong Zhang
2019 Neurocomputing  
We propose an AU synthesis framework that combines the well-known 3D Morphable Model (3DMM), which intrinsically disentangles expression parameters from other face attributes, with models that adversarially  ...  generate 3DMM expression parameters conditioned on given target AU labels, in contrast to the more conventional approach of generating facial images directly.  ...  Conclusions In this paper, we presented a framework for 3D facial action unit synthesis, which combines the advantages of both 3DMM and conditional generative adversarial models.  ... 
doi:10.1016/j.neucom.2019.05.003 fatcat:ds4uhemwivchljwy3mioui4ftq

Synthesis and Visualization of Photorealistic Textures for 3D Face Reconstruction of Prehistoric Human

Vladimir Kniaz, Vladimir Knyaz, Vladimir Mizginov
2020 Proceedings of the 30th International Conference on Computer Graphics and Machine Vision (GraphiCon 2020). Part 2  
We generate a dataset consisting of 3D models of human faces and skulls to train our 3D reconstruction model.  ...  Our framework leverages a joint face-skull model based on generative adversarial networks.  ...  Acknowledgements The reported study was funded by Russian Foundation for Basic Research (RFBR) according to the research project 17-29-04509.  ... 
doi:10.51130/graphicon-2020-2-4-15 fatcat:t3fqenpcojdpbknqaykvixwcsa

MACHINE LEARNING FOR APPROXIMATING UNKNOWN FACE

V. A. Knyaz, V. V. Kniaz, M. M. Novikov, R. M. Galeev
2020 The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences  
The skull 3D models used for appearance reconstruction are generated by the original photogrammetric system in automated mode.  ...  It exploits the generative adversarial learning for transition data from one modality (skull) to another modality (face) using digital skull 3D models and face 3D models.  ...  ACKNOWLEDGEMENTS The reported study was funded by Russian Foundation for Basic Research (RFBR) according to the research project 17-29-04509.  ... 
doi:10.5194/isprs-archives-xliii-b2-2020-857-2020 fatcat:s2rscqb525cfbivro3itakreia

A Novel Defensive Strategy for Facial Manipulation Detection Combining Bilateral Filtering and Joint Adversarial Training

Yifan Luo, Feng Ye, Bin Weng, Shan Du, Tianqiang Huang, Zhili Zhou
2021 Security and Communication Networks  
The introduction of joint adversarial training can train a model that defends against multiple adversarial attacks.  ...  The joint adversarial training starts from the training stage of the model, which mixes various adversarial examples and original examples to train the model.  ...  Specifically, adversarial examples are generated by model A to attack model B. If both are the same model, it is a white-box attack; otherwise, it is a black-box attack.  ... 
doi:10.1155/2021/4280328 fatcat:lijmatyrmfgpdbkz2wfsxh4654

UV-GAN: Adversarial Facial UV Map Completion for Pose-Invariant Face Recognition

Jiankang Deng, Shiyang Cheng, Niannan Xue, Yuxiang Zhou, Stefanos Zafeiriou
2018 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition  
Recently proposed robust 3D face alignment methods establish either dense or sparse correspondence between a 3D face model and a 2D facial image.  ...  To this end, we first gather complete UV maps by fitting a 3D Morphable Model (3DMM) to various multiview image and video datasets, as well as leveraging on a new 3D dataset with over 3,000 identities.  ...  Deng is supported by the President's Scholarship of Imperial College London. The work of S. Cheng is funded by the EPSRC project EP/N007743/1 (FACER2VM).  ... 
doi:10.1109/cvpr.2018.00741 dblp:conf/cvpr/DengCXZZ18 fatcat:nuz3vfdgabdk5eggpefhzgt6ky

3DFaceGAN: Adversarial Nets for 3D Face Representation, Generation, and Translation [article]

Stylianos Moschoglou, Stylianos Ploumpis, Mihalis Nicolaou, Athanasios Papaioannou, Stefanos Zafeiriou
2019 arXiv   pre-print
In this work, we present 3DFaceGAN, the first GAN tailored towards modeling the distribution of 3D facial surfaces, while retaining the high frequency details of 3D face shapes.  ...  Nevertheless, no GAN-based method has been proposed in the literature that can successfully represent, generate or translate 3D facial shapes (meshes).  ...  Acknowledgements Stylianos Moschoglou is supported by an EP-SRC DTA studentship from Imperial College London, Stylianos Ploumpis by the EPSRC Project EP/N007743/1 (FACER2VM), and Stefanos Zafeiriou by  ... 
arXiv:1905.00307v2 fatcat:5ggbmjhecrg2xazlupqbnx62ym

An Image-based Generator Architecture for Synthetic Image Refinement [article]

Alex Nasser
2021 arXiv   pre-print
Proposed are alternative generator architectures for Boundary Equilibrium Generative Adversarial Networks, motivated by Learning from Simulated and Unsupervised Images through Adversarial Training.  ...  It disentangles the need for a noise-based latent space. The generator will operate mainly as a refiner network to gain a photo-realistic presentation of the given synthetic images.  ...  Taking a more quantitative approach towards comparing the similarities between a 3D model and the generated faces, Betaface API, a facial recognition web service that supports face comparisons, has been  ... 
arXiv:2108.04957v1 fatcat:xqokp7ipdzbwrerhi3ibrdriyi

UV-GAN: Adversarial Facial UV Map Completion for Pose-invariant Face Recognition [article]

Jiankang Deng, Shiyang Cheng, Niannan Xue, Yuxiang Zhou, Stefanos Zafeiriou
2017 arXiv   pre-print
Recently proposed robust 3D face alignment methods establish either dense or sparse correspondence between a 3D face model and a 2D facial image.  ...  To this end, we first gather complete UV maps by fitting a 3D Morphable Model (3DMM) to various multiview image and video datasets, as well as leveraging on a new 3D dataset with over 3,000 identities.  ...  After fitting a 3DMM to the image, we retrieve a 3D face with an incomplete UV map. We learn a generative model to recover the self-occluded regions.  ... 
arXiv:1712.04695v1 fatcat:bejoef247vfqpdc7afgli5hciu

FaceDet3D: Facial Expressions with 3D Geometric Detail Prediction [article]

ShahRukh Athar, Albert Pumarola, Francesc Moreno-Noguer, Dimitris Samaras
2020 arXiv   pre-print
Facial Expressions induce a variety of high-level details on the 3D face geometry.  ...  The facial details are represented as a vertex displacement map and used then by a Neural Renderer to photo-realistically render novel images of any single image in any desired expression and view.  ...  Acknowledgements This work is supported in part by a Google Daydream Research award, by the Spanish government with projects HuMoUR TIN2017-90086-R and María de Maeztu Seal of Excellence MDM-2016-0656,  ... 
arXiv:2012.07999v3 fatcat:byexpip5ijhubhdcdn3jlziq4y

Semi-supervised 3D Face Representation Learning from Unconstrained Photo Collections

Zhongpai Gao, Juyong Zhang, Yudong Guo, Chao Ma, Guangtao Zhai, Xiaokang Yang
2020 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)  
We train our model in a semi-supervised manner with adversarial loss to exploit large amounts of unconstrained facial images.  ...  Recovering 3D geometry shape, albedo, and lighting from a single image is a typical ill-posed problem.  ...  This work was supported by the National Natural Science Foundation of China (61901259) and China Postdoctoral Science Foundation (BX2019208).  ... 
doi:10.1109/cvprw50498.2020.00182 dblp:conf/cvpr/GaoZG0ZY20 fatcat:yxtu23ltlnghfdtphnxzsjlira

A Survey on Face Data Augmentation [article]

Xiang Wang and Kai Wang and Shiguo Lian
2019 arXiv   pre-print
Among all these approaches, we put the emphasis on the deep learning-based works, especially the generative adversarial networks which have been recognized as more powerful and effective tools in recent  ...  used to enrich the training dataset.  ...  Generative-based Transformation The generative models provide a powerful tool to generate new data from modeled distribution by learning the data distribution of the training set.  ... 
arXiv:1904.11685v1 fatcat:phcwwc7gcfablgytt6itr6xade

Learning a High Fidelity Pose Invariant Model for High-resolution Face Frontalization [article]

Jie Cao, Yibo Hu, Hongwen Zhang, Ran He, Zhenan Sun
2018 arXiv   pre-print
Different from those reconstruction methods relying on 3D data, we also propose Adversarial Residual Dictionary Learning (ARDL) to supervise facial texture map recovering with only monocular images.  ...  We decompose the prerequisite of warping into dense correspondence field estimation and facial texture map recovering, which are both well addressed by deep networks.  ...  Recently, great breakthroughs have been made by the methods based on generative adversarial networks (GAN) [10] .  ... 
arXiv:1806.08472v2 fatcat:pflnq7o74be6vgxybzkzhb7y6a

3D Dense Geometry-Guided Facial Expression Synthesis by Adversarial Learning [article]

Rumeysa Bodur, Binod Bhattarai, Tae-Kyun Kim
2020 arXiv   pre-print
Manipulating facial expressions is a challenging task due to fine-grained shape changes produced by facial muscles and the lack of input-output pairs for supervised learning.  ...  To this end, we propose to use an off-the-shelf state-of-the-art 3D reconstruction model to estimate the depth and create a large-scale RGB-Depth dataset after a manual data clean-up process.  ...  Bodur is funded by the Turkish Ministry of National Education. This work is partly supported by EPSRC Programme Grant FACER2VM(EP/N007743/1).  ... 
arXiv:2009.14798v1 fatcat:p3vjpf4lpva7jir37o4qn6xqmq

3DFaceGAN: Adversarial Nets for 3D Face Representation, Generation, and Translation

Stylianos Moschoglou, Stylianos Ploumpis, Mihalis A. Nicolaou, Athanasios Papaioannou, Stefanos Zafeiriou
2020 International Journal of Computer Vision  
In this work, we present 3DFaceGAN, the first GAN tailored towards modeling the distribution of 3D facial surfaces, while retaining the high frequency details of 3D face shapes.  ...  Nevertheless, no GAN-based method has been proposed in the literature that can successfully represent, generate or translate 3D facial shapes (meshes).  ...  To view a copy of this licence, visit http://creativecomm ons.org/licenses/by/4.0/.  ... 
doi:10.1007/s11263-020-01329-8 fatcat:espnqtcoevcenbuinr7kmd3qqi

Adversarial Mask: Real-World Universal Adversarial Attack on Face Recognition Model [article]

Alon Zolfi and Shai Avidan and Yuval Elovici and Asaf Shabtai
2022 arXiv   pre-print
In addition, we validated our adversarial mask's effectiveness in real-world experiments (CCTV use case) by printing the adversarial pattern on a fabric face mask.  ...  In this paper, we propose Adversarial Mask, a physical universal adversarial perturbation (UAP) against state-of-the-art FR models that is applied on face masks in the form of a carefully crafted pattern  ...  Therefore, we use 3D face reconstruction to digitally apply a mask on a facial image. Feng et al.  ... 
arXiv:2111.10759v3 fatcat:s2nhmcpusnefzmty6rsfvfn2sa
« Previous Showing results 1 — 15 out of 4,207 results