How Far are We from Solving the 2D & 3D Face Alignment Problem? (and a Dataset of 230,000 3D Facial Landmarks)

Bulat, Adrian; Tzimiropoulos, Georgios

Abstract

This paper investigates how far a very deep neural network is from attaining close to saturating performance on existing 2D and 3D face alignment datasets. To this end, we make the following 5 contributions: (a) we construct, for the first time, a very strong baseline by combining a state-of-the-art architecture for landmark localization with a state-of-the-art residual block, train it on a very large yet synthetically expanded 2D facial landmark dataset and finally evaluate it on all other 2D facial landmark datasets. (b)We create a guided by 2D landmarks network which converts 2D landmark annotations to 3D and unifies all existing datasets, leading to the creation of LS3D-W, the largest and most challenging 3D facial landmark dataset to date (~230,000 images). (c) Following that, we train a neural network for 3D face alignment and evaluate it on the newly introduced LS3D-W. (d) We further look into the effect of all “traditional” factors affecting face alignment performance like large pose, initialization and resolution, and introduce a “new” one, namely the size of the network. (e) We show that both 2D and 3D face alignment networks achieve performance of remarkable accuracy which is probably close to saturating the datasets used. Training and testing code as well as the dataset can be downloaded from https: //www.adrianbulat.com/face-alignment/

Citation

Bulat, A., & Tzimiropoulos, G. (2017, October). How Far are We from Solving the 2D & 3D Face Alignment Problem? (and a Dataset of 230,000 3D Facial Landmarks). Presented at International Conference on Computer Vision (ICCV17), Venice, Italy

• Presentation Conference Type: Edited Proceedings
• Conference Name: International Conference on Computer Vision (ICCV17)
• Start Date: Oct 22, 2017
• End Date: Oct 29, 2017
• Acceptance Date: Jul 17, 2017
• Online Publication Date: Dec 25, 2017
• Publication Date: 2017-10
• Deposit Date: Aug 9, 2017
• Publicly Available Date: Oct 31, 2017
• Publisher: Institute of Electrical and Electronics Engineers
• Peer Reviewed: Peer Reviewed
• Volume: 2017-October
• Pages: 1021-1030
• Series Title: Proceedings (IEEE International Conference on Computer Vision)
• Series ISSN: 2380-7504
• Book Title: 2017 IEEE International Conference on Computer Vision (ICCV 2017)
• ISBN: 9781538610336
• DOI: https://doi.org/10.1109/ICCV.2017.116
• Public URL: https://nottingham-repository.worktribe.com/output/889423
• Publisher URL: http://ieeexplore.ieee.org/document/8237378/
• Related Public URLs: http://iccv2017.thecvf.com/
• Additional Information: © 2017 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works. Published in 2017 IEEE International Conference; on Computer Vision ISBN 9781538610329
• Contract Date: Aug 9, 2017