Filters








8,528 Hits in 8.3 sec

Learning to Regress 3D Face Shape and Expression from an Image without 3D Supervision [article]

Soubhik Sanyal, Timo Bolkart, Haiwen Feng, Michael J. Black
2019 arXiv   pre-print
To train a network without any 2D-to-3D supervision, we present RingNet, which learns to compute 3D face shape from a single image.  ...  The estimation of 3D face shape from a single image must be robust to variations in lighting, head pose, expression, facial hair, makeup, and occlusions.  ...  Osman for supplementary video, and S. Tang for useful discussions. Disclosure: Michael J. Black has received research gift funds from Intel, Nvidia, Adobe, Facebook, and Amazon.  ... 
arXiv:1905.06817v1 fatcat:hlpw2bjnl5dj3gk6zzueckhkdy

Learning to Regress 3D Face Shape and Expression From an Image Without 3D Supervision

Soubhik Sanyal, Timo Bolkart, Haiwen Feng, Michael J. Black
2019 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)  
Figure 1: Without 3D supervision, RingNet learns a mapping from the pixels of a single image to the 3D facial parameters of the FLAME model [21]. Top: Images are from the CelebA dataset [22].  ...  Bottom: estimated shape, pose and expression.  ...  Black has received research gift funds from Intel, Nvidia, Adobe, Facebook, and Amazon. He is a part-time employee of Amazon and has financial interests in Amazon and Meshcapade GmbH.  ... 
doi:10.1109/cvpr.2019.00795 dblp:conf/cvpr/SanyalBFB19 fatcat:bskyym6j2vewbmt372rxob5n5q

3D Face Reconstruction from A Single Image Assisted by 2D Face Images in the Wild [article]

Xiaoguang Tu, Jian Zhao, Zihang Jiang, Yao Luo, Mei Xie, Yang Zhao, Linxiao He, Zheng Ma, Jiashi Feng
2020 arXiv   pre-print
Recent methods typically aim to learn a CNN-based 3D face model that regresses coefficients of 3D Morphable Model (3DMM) from 2D images to render 3D face reconstruction or dense face alignment.  ...  Using these four self-supervision schemes, the 2DASL method significantly relieves demands on the the conventional paired 2D-to-3D annotations and gives much higher-quality 3D face models without requiring  ...  By leveraging these self-supervision derived from 2D face images without 3D annotations, our method could substantially improve the quality of learned 3D face regression model, even though there is lack  ... 
arXiv:1903.09359v2 fatcat:36qg3p4rine2fp2464j6jh2cni

Learning 3D Face Reconstruction with a Pose Guidance Network [article]

Pengpeng Liu, Xintong Han, Michael Lyu, Irwin King, Jia Xu
2020 arXiv   pre-print
3D face geometry from a single image.  ...  With our specially designed PGN, our model can learn from both faces with fully labeled 3D landmarks and unlimited unlabeled in-the-wild face images.  ...  CUHK 14210717 of the General Research Fund) and National Key Research and Development Program of China (No. 2018AAA0100204). We also thank Yao Feng, Feng Liu and Ayush Tewari for kind help.  ... 
arXiv:2010.04384v1 fatcat:j7xy3eks7rfsrbr3ildiqpjwea

Self-supervised Learning of Detailed 3D Face Reconstruction [article]

Yajing Chen, Fanzi Wu, Zeyu Wang, Yibing Song, Yonggen Ling, Linchao Bao
2019 arXiv   pre-print
In this paper, we present an end-to-end learning framework for detailed 3D face reconstruction from a single image.  ...  In the second stage, both the input image and the regressed texture of the coarse model are unwrapped into UV-space, and then sent through an image-toimage translation network to predict a displacement  ...  In this paper, we proposed a self-supervised model to learn 3D shape and texture in a coarse-to-fine procedure. In the coarse model, different from MoFA [6] and Genova et al.  ... 
arXiv:1910.11791v1 fatcat:p5ptxtiw5jflpfbwpelgtcmice

Lifting AutoEncoders: Unsupervised Learning of a Fully-Disentangled 3D Morphable Model Using Deep Non-Rigid Structure From Motion

Mihir Sahasrabudhe, Zhixin Shu, Edward Bartrum, Riza Alp Guler, Dimitris Samaras, Iasonas Kokkinos
2019 2019 IEEE/CVF International Conference on Computer Vision Workshop (ICCVW)  
Figure 1 : We introduce Lifting AutoEncoders, a deep generative model of 3D shape variability that is learned from an unstructured photo collection without supervision.  ...  We further use weak supervision to disentangle the non-rigid shape variability of human faces into identity and expression.  ...  This work was supported by a gift from Adobe, NSF grants CNS-1718014 and DMS 1737876, the Partner University Fund, and the SUNY2020 Infrastructure Transportation Security Center.  ... 
doi:10.1109/iccvw.2019.00500 dblp:conf/iccvw/SahasrabudheSBG19 fatcat:f3dv3unkqnhvphph76laefew54

3D Face From X: Learning Face Shape from Diverse Sources [article]

Yudong Guo, Lin Cai, Juyong Zhang
2021 arXiv   pre-print
We present a novel method to jointly learn a 3D face parametric model and 3D face reconstruction from diverse sources.  ...  Besides scanned face data and face images, we also utilize a large number of RGB-D images captured with an iPhone X to bridge the gap between the two sources.  ...  [11] presents a learning-based regression approach to fit a generic identity and expression model to an RGB face video on the fly.  ... 
arXiv:1808.05323v3 fatcat:2ivplklvcnfkradpqjucqqf5j4

Accurate 3D Face Reconstruction with Weakly-Supervised Learning: From Single Image to Image Set [article]

Yu Deng, Jiaolong Yang, Sicheng Xu, Dong Chen, Yunde Jia, Xin Tong
2020 arXiv   pre-print
information for supervision, and 2) performs multi-image face reconstruction by exploiting complementary information from different images for shape aggregation.  ...  , whereas face images with ground-truth 3D face shapes are scarce.  ...  Nor does it take full advantage of pose differences to improve the shape prediction. In this paper, we propose to learn 3D face aggregation from multiple images, also in an unsupervised fashion.  ... 
arXiv:1903.08527v2 fatcat:nwscyixadfeqlktsv4vjyiqzpa

From 2D Images to 3D Model:Weakly Supervised Multi-View Face Reconstruction with Deep Fusion [article]

Weiguang Zhao and Chaolong Yang and Jianan Ye and Yuyao Yan and Xi Yang and Kaizhu Huang
2022 arXiv   pre-print
We consider the problem of Multi-view 3D Face Reconstruction (MVR) with weakly supervised learning that leverages a limited number of 2D face images (e.g. 3) to generate a high-quality 3D face model with  ...  Without 3D landmarks annotation, DF-MVR achieves 5.2% and 3.0% RMSE improvements over the existing best weakly supervised MVRs respectively on Pixel-Face and Bosphorus datasets; with 3D landmarks annotation  ...  The image with high confidence is used to regress shape coefficients, and the rest images will be used to regress coefficients such as expression and texture. [11] adopts the concept of geometry consistency  ... 
arXiv:2204.03842v2 fatcat:rzvbjdew4vffzolfwxus7y35gy

Accurate 3D Face Reconstruction With Weakly-Supervised Learning: From Single Image to Image Set

Yu Deng, Jiaolong Yang, Sicheng Xu, Dong Chen, Yunde Jia, Xin Tong
2019 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)  
information for supervision, and 2) performs multi-image face reconstruction by exploiting complementary information from different images for shape aggregation.  ...  However, training deep neural networks typically requires a large volume of data, whereas face images with ground-truth 3D face shapes are scarce.  ...  Nor does it take full advantage of pose differences to improve the shape prediction. In this paper, we propose to learn 3D face aggregation from multiple images, also in an unsupervised fashion.  ... 
doi:10.1109/cvprw.2019.00038 dblp:conf/cvpr/DengYX0JT19 fatcat:xcfulxaqfjafjlwoasjlcxw5r4

Lifting AutoEncoders: Unsupervised Learning of a Fully-Disentangled 3D Morphable Model using Deep Non-Rigid Structure from Motion [article]

Mihir Sahasrabudhe, Zhixin Shu, Edward Bartrum, Riza Alp Guler, Dimitris Samaras, Iasonas Kokkinos
2019 arXiv   pre-print
We further use weak supervision to disentangle the non-rigid shape variability of human faces into identity and expression.  ...  We bring together ideas from non-rigid structure from motion, image formation, and morphable models to learn a controllable, geometric model of 3D categories in an entirely unsupervised manner from an  ...  For shape modeling we use sources of weak supervision to factor the shape variability into 3D pose, and non-rigid identity and expression, allowing us to control the expression or identity of a face by  ... 
arXiv:1904.11960v1 fatcat:ohpo2bu3rbe2hm5dwq4k55cnv4

A Self-Supervised Bootstrap Method for Single-Image 3D Face Reconstruction [article]

Yifan Xing, Rahul Tewari, Paulo R. S. Mendonca
2018 arXiv   pre-print
3D reconstruction of faces in near-frontal images without ground-truth 3D shape; (ii) application of a rigid-body transformation to the reconstructed face model; (iii) rendering of the face model from  ...  State-of-the-art methods for 3D reconstruction of faces from a single image require 2D-3D pairs of ground-truth data for supervision.  ...  of learning without forgetting from [41] .  ... 
arXiv:1812.05806v2 fatcat:z7rvid4qrnaktfgch25qhirlma

FML: Face Model Learning From Videos

Ayush Tewari, Florian Bernard, Pablo Garrido, Gaurav Bharaj, Mohamed Elgharib, Hans-Peter Seidel, Patrick Perez, Michael Zollhofer, Christian Theobalt
2019 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)  
In contrast, we propose multi-frame video-based self-supervised training of a deep network that (i) learns a face identity model both in shape and appearance while (ii) jointly learning to reconstruct  ...  We propose multi-frame self-supervised training of a deep network based on in-the-wild video data for jointly learning a face model and 3D face reconstruction.  ...  Acknowledgements We thank True-VisionSolutions Pty Ltd for providing the 2D face tracker, and the authors of [12, 48, 52, 62] for the comparisons.  ... 
doi:10.1109/cvpr.2019.01107 dblp:conf/cvpr/TewariB0BESPZT19 fatcat:6gf5b75bkzbldhzbyqnun4okzm

FML: Face Model Learning from Videos [article]

Ayush Tewari, Florian Bernard, Pablo Garrido, Gaurav Bharaj, Mohamed Elgharib, Hans-Peter Seidel, Patrick Pérez, Michael Zollhöfer, Christian Theobalt
2019 arXiv   pre-print
In contrast, we propose multi-frame video-based self-supervised training of a deep network that (i) learns a face identity model both in shape and appearance while (ii) jointly learning to reconstruct  ...  Our face model is learned using only corpora of in-the-wild video clips collected from the Internet. This virtually endless source of training data enables learning of a highly general 3D face model.  ...  Acknowledgements: We thank True-VisionSolutions Pty Ltd for providing the 2D face tracker, and the authors of [12, 48, 52, 62] for the comparisons.  ... 
arXiv:1812.07603v2 fatcat:mdnemyu7xjf5lbhszg5i3e53fu

Synergy between 3DMM and 3D Landmarks for Accurate 3D Facial Geometry [article]

Cho-Ying Wu, Qiangeng Xu, Ulrich Neumann
2021 arXiv   pre-print
This work studies learning from a synergy process of 3D Morphable Models (3DMM) and 3D facial landmarks to predict complete 3D facial geometry, including 3D alignment, face orientation, and 3D face modeling  ...  Our synergy process leverages a representation cycle for 3DMM parameters and 3D landmarks. 3D landmarks can be extracted and refined from face meshes built by 3DMM parameters.  ...  Acknowledgement We sincerely thank Jingjing Zheng, Jim Thomas, and Cheng-Hao Kuo for their detailed feedback on this paper.  ... 
arXiv:2110.09772v2 fatcat:4teppdxktvffxf3vj5cw4bjnqq
« Previous Showing results 1 — 15 out of 8,528 results