Real-Time 3D Face Alignment Using an Encoder-Decoder Network With an Efficient Deconvolution Layer
- 22 October 2020
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Signal Processing Letters
- Vol. 27 (10709908) , 1944-1948
Abstract
In the field of 3D face alignment, most researchers have focused on improving the prediction accuracy of algorithms and ignored the portability for practical applications. To this end, this study presents a real-time 3D face-alignment method that uses an encoder–decoder network with an efficient deconvolution layer. The fusion of the encoding and decoding feature adds more abundant features to this network. An efficient deconvolution layer at the decoding stage applies the L1 norm to select useful features and generate abundant ones through linear operations. Experimental results using the standard AFLW2000-3D and AFLW-LFPA datasets show that our algorithm has low prediction errors with real-time applicability.Keywords
Funding Information
- National Natural Science Foundation of China (61901436)
This publication has 21 references indexed in Scilit:
- Stacked Hourglass Networks for Human Pose Estimation2016
- Large-Pose Face Alignment via CNN-Based Dense 3D Model Fitting2016
- Deep Learning Face Attributes in the Wild2015
- Pose-Invariant 3D Face Alignment2015
- U-Net: Convolutional Networks for Biomedical Image Segmentation2015
- Robust Face Recognition via Adaptive Sparse RepresentationIEEE Transactions on Cybernetics, 2014
- Face Alignment by Explicit Shape RegressionInternational Journal of Computer Vision, 2013
- Supervised Descent Method and Its Applications to Face Alignment2013
- 3D Face Reconstruction from a Single Image Using a Single Reference Face ShapeIEEE Transactions on Pattern Analysis and Machine Intelligence, 2010
- A morphable model for the synthesis of 3D faces1999