Filters








12,070 Hits in 4.8 sec

Synthesizing Normalized Faces from Facial Identity Features [article]

Forrester Cole, David Belanger, Dilip Krishnan, Aaron Sarna, Inbar Mosseri, William T. Freeman
2017 arXiv   pre-print
This is achieved by learning to generate facial landmarks and textures from features extracted from a facial-recognition network.  ...  We present a method for synthesizing a frontal, neutral-expression image of a person's face given an input face photograph.  ...  There is no obvious way to reverse the embedding and produce an image of a face from a given feature vector. We present a method for mapping from facial identity features back to images of faces.  ... 
arXiv:1701.04851v4 fatcat:rok4xq6jbrcyvmfs4zls5axmvq

Synthesizing Normalized Faces from Facial Identity Features

Forrester Cole, David Belanger, Dilip Krishnan, Aaron Sarna, Inbar Mosseri, William T. Freeman
2017 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)  
This is achieved by learning to generate facial landmarks and textures from features extracted from a facial-recognition network.  ...  We present a method for synthesizing a frontal, neutralexpression image of a person's face given an input face photograph.  ...  There is no obvious way to reverse the embedding and produce an image of a face from a given feature vector. We present a method for mapping from facial identity features back to images of faces.  ... 
doi:10.1109/cvpr.2017.361 dblp:conf/cvpr/ColeBKSMF17 fatcat:ccou3dhcfje5rbkeif2mbbv2ia

LandmarkGAN: Synthesizing Faces from Landmarks [article]

Pu Sun, Yuezun Li, Honggang Qi, Siwei Lyu
2021 arXiv   pre-print
Face synthesis is an important problem in computer vision with many applications. In this work, we describe a new method, namely LandmarkGAN, to synthesize faces based on facial landmarks as input.  ...  Facial landmarks are a natural, intuitive, and effective representation for facial expressions and orientations, which are independent from the target's texture or color and background scene.  ...  The work [20] synthesizes faces from facial landmarks.  ... 
arXiv:2011.00269v2 fatcat:agjvikwncrdfpagle2jsbxislq

Adaptive 3D Model-Based Facial Expression Synthesis and Pose Frontalization

Yu-Jin Hong, Sung Eun Choi, Gi Pyo Nam, Heeseung Choi, Junghyun Cho, Ig-Jae Kim
2020 Sensors  
However, owing to complex and subtle facial muscle movements, facial expression modeling from images with face poses is difficult to achieve.  ...  To do so, we constructed a wrinkle table of various facial expressions from 400 people.  ...  If the normal direction is opposite, the vertex is invisible at that position and we exclude the point from the contour features.  ... 
doi:10.3390/s20092578 pmid:32369980 pmcid:PMC7248866 fatcat:zgb7tbjk2jd2jodkaacwtwwxie

One-shot Face Reenactment Using Appearance Adaptive Normalization [article]

Guangming Yao, Yi Yuan, Tianjia Shao, Shuang Li, Shanqi Liu, Yong Liu, Mengmeng Wang, Kun Zhou
2021 arXiv   pre-print
The core of our network is a novel mechanism called appearance adaptive normalization, which can effectively integrate the appearance information from the input image into our face generator by modulating  ...  the feature maps of the generator using the learned adaptive parameters.  ...  In general, the appearance adaptive normalization can effectively integrate the specific appearance information from the source image into the synthesized image, by modulating the feature maps of the face  ... 
arXiv:2102.03984v3 fatcat:ocphlr5whbbnrbcztpwmk5ofxe

Heterogeneous Face Recognition via Face Synthesis with Identity-Attribute Disentanglement

Ziming Yang, Jian Liang, Chaoyou Fu, Mandi Luo, Xiao-Yu Zhang
2022 IEEE Transactions on Information Forensics and Security  
To address these challenges, we propose a new HFR method from the perspective of heterogeneous data augmentation, named Face Synthesis with Identity-Attribute Disentanglement (FSIAD).  ...  Firstly, the identity-attribute disentanglement (IAD) decouples face images into identity-related representations and identity-unrelated representations (called attributes), and then decreases the correlation  ...  synthesize multimodal faces from the predetermined visual descriptions of facial attributes.  ... 
doi:10.1109/tifs.2022.3160595 fatcat:zwwzaxz57vbxfitmp6ebtehoya

Multi-view Face Recognition via Well-advised Pose Normalization Network

Xiaohu Shao, Xiangdong Zhou, Zhenghao Li, Yu Shi
2020 IEEE Access  
However, facial feature analysis and identity discrimination often suffer from failure frontalization results because of monotonous single-domain training and unpredictable input profile faces.  ...  Meanwhile, the proposed method encourages intra-class compactness and interclass separability between facial features by introducing quality-aware feature fusion.  ...  POSE NORMALIZATION NETWORK Given a profile face image x with an identity label l id , the goal of the pose normalization network is to synthesize a frontal face x with realistic facial textures while preserving  ... 
doi:10.1109/access.2020.2983459 fatcat:y4ozlaqrbrdvbgvnyckcwy3zli

Synthesizing Coupled 3D Face Modalities by Trunk-Branch Generative Adversarial Networks [article]

Baris Gecer, Alexander Lattas, Stylianos Ploumpis, Jiankang Deng, Athanasios Papaioannou, Stylianos Moschoglou, Stefanos Zafeiriou
2020 arXiv   pre-print
Nevertheless, these models cannot represent faithfully either the facial texture or the normals of the face, which are very crucial for photo-realistic face synthesis.  ...  Generally, research on 3D face generation revolves around linear statistical models of the facial surface.  ...  All of these studies show the significance of photorealistic and identity-generic face synthesization for the next generation of facial recognition algorithms.  ... 
arXiv:1909.02215v2 fatcat:atgjbzfgsjeplimnmgo5h5ck7a

Generating Synthetic Disguised Faces with Cycle-Consistency Loss and an Automated Filtering Algorithm

Mobeen Ahmad, Usman Cheema, Muhammad Abdullah, Seungbin Moon, Dongil Han
2021 Mathematics  
Cycle-consistency loss is used to generate facial images with disguises, e.g., fake beards, makeup, and glasses, from normal face images.  ...  Additionally, an automated filtering scheme is presented for automated data filtering from the synthesized faces.  ...  The proposed methodology synthesizes disguised facial images using nondisguised facial images, while preserving the identity and the representative features in the synthesized image.  ... 
doi:10.3390/math10010004 fatcat:nnv2j5qakjev3iormdz6j36n54

LEED: Label-Free Expression Editing via Disentanglement [article]

Rongliang Wu, Shijian Lu
2020 arXiv   pre-print
The idea is to disentangle the identity and expression of a facial image in the expression manifold, where the neutral face captures the identity attribute and the displacement between the neutral image  ...  siamese loss that aims to enhance the expression similarity between the synthesized image and the reference image.  ...  The FID scores are calculated between the final average pooling features of a pre-trained inception model [46] of the real faces and the synthesized faces, and the SSIM is computed over synthesized expressions  ... 
arXiv:2007.08971v1 fatcat:wtbaxrdenndxxgahoe7tsn75ay

An Efficient Integration of Disentangled Attended Expression and Identity FeaturesFor Facial Expression Transfer andSynthesis [article]

Kamran Ali, Charles E. Hughes
2020 arXiv   pre-print
In this paper, we present an Attention-based Identity Preserving Generative Adversarial Network (AIP-GAN) to overcome the identity leakage problem from a source image to a generated face image, an issue  ...  Similarly, the disentangled expression-agnostic identity features are extracted from the input target image by inferring its combined intrinsic-shape and appearance image employing our self-supervised  ...  identity information of an unseen face image during inference since they fail to disentangle expression features from identity representation.  ... 
arXiv:2005.00499v1 fatcat:xw2wapx76vbcbguedh34helqyu

Facial expression transfer method based on frequency analysis

Wei Wei, Chunna Tian, Stephen John Maybank, Yanning Zhang
2016 Pattern Recognition  
We locate the facial features automatically and describe the shape deformations between a neutral expression and non-neutral expressions.  ...  The resulting synthesized image preserves both the facial appearance of the target subject and the expression details of the source subject.  ...  (d) is the facial texture extracted from (b). (e) is the warped face of (a) to the shape of (e). (f) is the warped face of (c) to the shape of (e).  ... 
doi:10.1016/j.patcog.2015.08.004 fatcat:gjuziqll25fvphgujajtglmmkm

Learning Flow-based Feature Warping for Face Frontalization with Illumination Inconsistent Supervision [article]

Yuxiang Wei, Ming Liu, Haolin Wang, Ruifeng Zhu, Guosheng Hu, Wangmeng Zuo
2020 arXiv   pre-print
We propose a novel Flow-based Feature Warping Model (FFWM) which can learn to synthesize photo-realistic and illumination preserving frontal images with illumination inconsistent supervision.  ...  Moreover, a Warp Attention Module (WAM) is introduced to reduce the pose discrepancy in the feature level, and hence to synthesize frontal images more effectively and preserve more details of profile images  ...  Each testing identity has one gallery image with frontal view and normal illumination from the first appearance. LFW [10] contains 13,233 face images collected in unconstrained environment.  ... 
arXiv:2008.06843v2 fatcat:jelehk45mja2zcqitpnvzrnctm

Learn to synthesize and synthesize to learn

Behzad Bozorgtabar, Mohammad Saeed Rad, Hazım Kemal Ekenel, Jean-Philippe Thiran
2019 Computer Vision and Image Understanding  
To overcome these shortcomings, we propose attribute guided face image generation method using a single model, which is capable to synthesize multiple photo-realistic face images conditioned on the attributes  ...  Finally, we demonstrate that generated facial images can be used for synthetic data augmentation, and improve the performance of the classifier used for facial expression recognition.  ...  Synthesizing photo-realistic facial images has applications in human-computer interactions, facial animation and more importantly in facial identity or expression recognition.  ... 
doi:10.1016/j.cviu.2019.04.010 fatcat:z6h6o2foqrfapchl74bgf7ssbq

Learn to synthesize and synthesize to learn [article]

Behzad Bozorgtabar, Mohammad Saeed Rad, Hazım Kemal Ekenel and Jean-Philippe Thiran
2019 arXiv   pre-print
To overcome these shortcomings, we propose attribute guided face image generation method using a single model, which is capable to synthesize multiple photo-realistic face images conditioned on the attributes  ...  Finally, we demonstrate that generated facial images can be used for synthetic data augmentation, and improve the performance of the classifier used for facial expression recognition.  ...  Synthesizing photo-realistic facial images has applications in human-computer interactions, facial animation and more importantly in facial identity or expression recognition.  ... 
arXiv:1905.00286v1 fatcat:xjpgodrpjjcidm7yugogsd7av4
« Previous Showing results 1 — 15 out of 12,070 results