GeoConv: Geodesic Guided Convolution for Facial Action Unit Recognition

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Chen, Yuedong, Song, Guoxian, Shao, Zhiwen, Cai, Jianfei, Cham, Tat-Jen, Zheng, Jianming
Natura: Preprint
Pubblicazione: 2020
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866908488211365888
author Chen, Yuedong
Song, Guoxian
Shao, Zhiwen
Cai, Jianfei
Cham, Tat-Jen
Zheng, Jianming
author_facet Chen, Yuedong
Song, Guoxian
Shao, Zhiwen
Cai, Jianfei
Cham, Tat-Jen
Zheng, Jianming
contents Automatic facial action unit (AU) recognition has attracted great attention but still remains a challenging task, as subtle changes of local facial muscles are difficult to thoroughly capture. Most existing AU recognition approaches leverage geometry information in a straightforward 2D or 3D manner, which either ignore 3D manifold information or suffer from high computational costs. In this paper, we propose a novel geodesic guided convolution (GeoConv) for AU recognition by embedding 3D manifold information into 2D convolutions. Specifically, the kernel of GeoConv is weighted by our introduced geodesic weights, which are negatively correlated to geodesic distances on a coarsely reconstructed 3D face model. Moreover, based on GeoConv, we further develop an end-to-end trainable framework named GeoCNN for AU recognition. Extensive experiments on BP4D and DISFA benchmarks show that our approach significantly outperforms the state-of-the-art AU recognition methods.
format Preprint
id arxiv_https___arxiv_org_abs_2003_03055
institution arXiv
publishDate 2020
record_format arxiv
spellingShingle GeoConv: Geodesic Guided Convolution for Facial Action Unit Recognition
Chen, Yuedong
Song, Guoxian
Shao, Zhiwen
Cai, Jianfei
Cham, Tat-Jen
Zheng, Jianming
Computer Vision and Pattern Recognition
Automatic facial action unit (AU) recognition has attracted great attention but still remains a challenging task, as subtle changes of local facial muscles are difficult to thoroughly capture. Most existing AU recognition approaches leverage geometry information in a straightforward 2D or 3D manner, which either ignore 3D manifold information or suffer from high computational costs. In this paper, we propose a novel geodesic guided convolution (GeoConv) for AU recognition by embedding 3D manifold information into 2D convolutions. Specifically, the kernel of GeoConv is weighted by our introduced geodesic weights, which are negatively correlated to geodesic distances on a coarsely reconstructed 3D face model. Moreover, based on GeoConv, we further develop an end-to-end trainable framework named GeoCNN for AU recognition. Extensive experiments on BP4D and DISFA benchmarks show that our approach significantly outperforms the state-of-the-art AU recognition methods.
title GeoConv: Geodesic Guided Convolution for Facial Action Unit Recognition
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2003.03055