Deep learning the dynamic appearance and shape of facial action units

Spontaneous facial expression recognition under uncontrolled conditions is a hard task. It depends on multiple factors including shape, appearance and dynamics of the facial features, all of which are adversely affected by environmental noise and low intensity signals typical of such conditions. In...

Full description

Bibliographic Details
Main Authors: Jaiswal, Shashank, Valstar, Michel F.
Format: Conference or Workshop Item
Published: 2016
Online Access:https://eprints.nottingham.ac.uk/31301/
_version_ 1848794171839610880
author Jaiswal, Shashank
Valstar, Michel F.
author_facet Jaiswal, Shashank
Valstar, Michel F.
author_sort Jaiswal, Shashank
building Nottingham Research Data Repository
collection Online Access
description Spontaneous facial expression recognition under uncontrolled conditions is a hard task. It depends on multiple factors including shape, appearance and dynamics of the facial features, all of which are adversely affected by environmental noise and low intensity signals typical of such conditions. In this work, we present a novel approach to Facial Action Unit detection using a combination of Convolutional and Bi-directional Long Short-Term Memory Neural Networks (CNN-BLSTM), which jointly learns shape, appearance and dynamics in a deep learning manner. In addition, we introduce a novel way to encode shape features using binary image masks computed from the locations of facial landmarks. We show that the combination of dynamic CNN features and Bi-directional Long Short-Term Memory excels at modelling the temporal information. We thoroughly evaluate the contributions of each component in our system and show that it achieves state-of-the-art performance on the FERA-2015 Challenge dataset.
first_indexed 2025-11-14T19:11:57Z
format Conference or Workshop Item
id nottingham-31301
institution University of Nottingham Malaysia Campus
institution_category Local University
last_indexed 2025-11-14T19:11:57Z
publishDate 2016
recordtype eprints
repository_type Digital Repository
spelling nottingham-313012020-05-04T20:05:18Z https://eprints.nottingham.ac.uk/31301/ Deep learning the dynamic appearance and shape of facial action units Jaiswal, Shashank Valstar, Michel F. Spontaneous facial expression recognition under uncontrolled conditions is a hard task. It depends on multiple factors including shape, appearance and dynamics of the facial features, all of which are adversely affected by environmental noise and low intensity signals typical of such conditions. In this work, we present a novel approach to Facial Action Unit detection using a combination of Convolutional and Bi-directional Long Short-Term Memory Neural Networks (CNN-BLSTM), which jointly learns shape, appearance and dynamics in a deep learning manner. In addition, we introduce a novel way to encode shape features using binary image masks computed from the locations of facial landmarks. We show that the combination of dynamic CNN features and Bi-directional Long Short-Term Memory excels at modelling the temporal information. We thoroughly evaluate the contributions of each component in our system and show that it achieves state-of-the-art performance on the FERA-2015 Challenge dataset. 2016 Conference or Workshop Item PeerReviewed Jaiswal, Shashank and Valstar, Michel F. (2016) Deep learning the dynamic appearance and shape of facial action units. In: Winter Conference on Applications of Computer Vision (WACV), 7-9 March 2016, Lake Placid, USA. (In Press)
spellingShingle Jaiswal, Shashank
Valstar, Michel F.
Deep learning the dynamic appearance and shape of facial action units
title Deep learning the dynamic appearance and shape of facial action units
title_full Deep learning the dynamic appearance and shape of facial action units
title_fullStr Deep learning the dynamic appearance and shape of facial action units
title_full_unstemmed Deep learning the dynamic appearance and shape of facial action units
title_short Deep learning the dynamic appearance and shape of facial action units
title_sort deep learning the dynamic appearance and shape of facial action units
url https://eprints.nottingham.ac.uk/31301/