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

Bouafif, Oussema; Khomutenko, Bogdan; Daoudi, Mohamed

Document type :

Autre communication scientifique (congrès sans actes - poster - séminaire...): Communication dans un congrès avec actes

Title :

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

Author(s) :

Bouafif, Oussema [Auteur]
Centre de Recherche en Informatique, Signal et Automatique de Lille - UMR 9189 [CRIStAL]
Khomutenko, Bogdan [Auteur]
Daoudi, Mohamed [Auteur]

Ecole nationale supérieure Mines-Télécom Lille Douai [IMT Lille Douai]
Centre de Recherche en Informatique, Signal et Automatique de Lille - UMR 9189 [CRIStAL]

Conference title :

15th International Conference on Computer Vision, Theory and Applications.

City :

Valletta

Country :

Malte

Start date of the conference :

2020-02-27

English keyword(s) :

3D Head Reconstruction
Face Reconstruction
Monocular Reconstruction
Facial Surface Normals
Deep Learning
Synthetic Data

HAL domain(s) :

Informatique [cs]/Vision par ordinateur et reconnaissance de formes [cs.CV]

English abstract : [en]

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 ...
Show more >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.Show less >

Language :

Anglais

Peer reviewed article :

Oui

Audience :

Internationale

Popular science :

Non

Collections :

Centre de Recherche en Informatique, Signal et Automatique de Lille (CRIStAL) - UMR 9189

Source :

Harvested from HAL

Files

https://hal.archives-ouvertes.fr/hal-02928494/document
Open access
Access the document

https://hal.archives-ouvertes.fr/hal-02928494/document
Open access
Access the document

https://hal.archives-ouvertes.fr/hal-02928494/document
Open access
Access the document

document
Open access
Access the document

VISAPP2020-2.pdf
Open access
Access the document

Monocular 3D Head Reconstruction via ... BibTeX CSV Excel RIS

Files

Monocular 3D Head Reconstruction via ...

BibTeX

CSV

Excel

RIS