Publication

Detailed 3D face reconstruction from a single RGB image

Conference Article

Conference

International Conference in Central Europe on Computer Graphics, Visualization and Computer Vision (WSCG)

Edition

27th

Doc link

http://wscg.zcu.cz/DL/wscg_DL.htm

Document original web page

Abstract

This paper introduces a method to obtain a detailed 3D reconstruction of facial skin from a single RGB image. To this end, we propose the exclusive use of an input image without requiring any information about the observed material nor training data to model the wrinkle properties. They are detected and characterized directly from the image via a simple and effective parametric model, determining several features such as location, orientation, width, and height. With these ingredients, we propose to minimize a photometric error to retrieve the final detailed 3D map, which is initialized by current techniques based on deep learning. In contrast with other approaches, we only require estimating a depth parameter, making our approach fast and intuitive. Extensive experimental evaluation is presented in a wide variety of synthetic and real images, including different skin properties and facial expressions. In all cases, our method outperforms the current approaches regarding 3D reconstruction accuracy, providing striking results for both large and fine wrinkles.

Categories

computer vision.

Author keywords

3D Wrinkle Reconstruction, Face Analysis, Optimization.

Scientific reference

G. Rotger, F. Moreno-Noguer, F. Lumbreras and A. Agudo. Detailed 3D face reconstruction from a single RGB image, 27th International Conference in Central Europe on Computer Graphics, Visualization and Computer Vision, 2019, Plzen, Czech Republic, to appear.