End to End Face Reconstruction via Differentiable PnP

Abstract

This is a challenge report of the ECCV 2022 WCPA Challenge, Face Reconstruction Track. Inside this report is a brief explanation of how we accomplish this challenge. We design a two-branch network to accomplish this task, whose roles are Face Reconstruction and Face Landmark Detection. The former outputs canonical 3D face coordinates. The latter outputs pixel coordinates, i.e. 2D mapping of 3D coordinates with head pose and perspective projection. In addition, we utilize a differentiable PnP (Perspective-n-Points) layer to finetune the outputs of the two branch. Our method achieves very competitive quantitative results on the MVP-Human dataset and wins a $3^{rd}$ prize in the challenge.

Accepted by ECCV2022 workshop

References in corpus (5)

• An Intriguing Failing of Convolutional Neural Networks and the CoordConv Solution
• 6D Rotation Representation For Unconstrained Head Pose Estimation
• Joint 3D Face Reconstruction and Dense Alignment with Position Map Regression Network
• Learning an Animatable Detailed 3D Face Model from In-The-Wild Images
• 3D face reconstruction with dense landmarks