Pix2face: Direct 3D Face Model Estimation
Daniel Crispell, Maxim Bazik
TL;DR
This paper introduces Pix2face, an automatic neural network-based method for estimating 3D face shape and pose from 2D images, utilizing a modified U-Net architecture and direct 3D Morphable Model parameter estimation.
Contribution
It presents a novel approach combining dense 3D landmark estimation with direct 3DMM parameter prediction in a single framework.
Findings
Accurate 3D face landmarks in unconstrained videos
Effective joint estimation of face geometry and pose
Qualitative and quantitative validation results
Abstract
An efficient, fully automatic method for 3D face shape and pose estimation in unconstrained 2D imagery is presented. The proposed method jointly estimates a dense set of 3D landmarks and facial geometry using a single pass of a modified version of the popular "U-Net" neural network architecture. Additionally, we propose a method for directly estimating a set of 3D Morphable Model (3DMM) parameters, using the estimated 3D landmarks and geometry as constraints in a simple linear system. Qualitative modeling results are presented, as well as quantitative evaluation of predicted 3D face landmarks in unconstrained video sequences.
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.