Monocular 3D Head Reconstruction via Prediction and Integration of Normal Vector Field

Oussema Bouafif, Oussema Bouafif, Bogdan Khomutenko, Mohamed Daoudi

2020

Abstract

Reconstructing the geometric structure of a face from a single input image is a challenging active research area in computer vision. In this paper, we present a novel method for reconstructing 3D heads from an input image using a hybrid approach based on learning and geometric techniques. We introduce a deep neural network trained on synthetic data only, which predicts the map of normal vectors of the face surface from a single photo. Afterward, using the network output we recover the 3D facial geometry by means of weighted least squares. Through qualitative and quantitative evaluation tests, we show the accuracy and robustness of our proposed method. Our method does not require accurate alignment due to the image-to-image translation network and also successfully recovers 3D geometry for real images, despite the fact that the model was trained only on synthetic data.

Paper Citation

in Harvard Style

Bouafif O., Khomutenko B. and Daoudi M. (2020). Monocular 3D Head Reconstruction via Prediction and Integration of Normal Vector Field. In Proceedings of the 15th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2020) - Volume 5: VISAPP; ISBN 978-989-758-402-2, SciTePress, pages 359-369. DOI: 10.5220/0008961703590369

in Bibtex Style

@conference{visapp20,
author={Oussema Bouafif and Bogdan Khomutenko and Mohamed Daoudi},
title={Monocular 3D Head Reconstruction via Prediction and Integration of Normal Vector Field},
booktitle={Proceedings of the 15th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2020) - Volume 5: VISAPP},
year={2020},
pages={359-369},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0008961703590369},
isbn={978-989-758-402-2},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 15th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2020) - Volume 5: VISAPP
TI - Monocular 3D Head Reconstruction via Prediction and Integration of Normal Vector Field
SN - 978-989-758-402-2
AU - Bouafif O.
AU - Khomutenko B.
AU - Daoudi M.
PY - 2020
SP - 359
EP - 369
DO - 10.5220/0008961703590369
PB - SciTePress