Use this URL to cite or link to this record in EThOS:
Title: Deep learning for facial emotion recognition
Author: Ruiz-Garcia, Ariel
ISNI:       0000 0004 7962 2928
Awarding Body: Coventry University
Current Institution: Coventry University
Date of Award: 2018
Availability of Full Text:
Access from EThOS:
Access from Institution:
The ability to perceive and interpret human emotions is an essential as-pect of daily life. The recent success of deep learning (DL) has resulted in the ability to utilize automated emotion recognition by classifying af-fective modalities into a given emotional state. Accordingly, DL has set several state-of-the-art benchmarks on static affective corpora collected in controlled environments. Yet, one of the main limitations of DL based intelligent systems is their inability to generalize on data with nonuniform conditions. For instance, when dealing with images in a real life scenario, where extraneous variables such as natural or artificial lighting are sub-ject to constant change, the resulting changes in the data distribution commonly lead to poor classification performance. These and other con-straints, such as: lack of realistic data, changes in facial pose, and high data complexity and dimensionality increase the difficulty of designing DL models for emotion recognition in unconstrained environments. This thesis investigates the development of deep artificial neural net-work learning algorithms for emotion recognition with specific attention to illumination and facial pose invariance. Moreover, this research looks at the development of illumination and rotation invariant face detection architectures based on deep reinforcement learning. The contributions and novelty of this thesis are presented in the form of several deep learning pose and illumination invariant architectures that offer state-of-the-art classification performance on data with nonuniform conditions. Furthermore, a novel deep reinforcement learning architecture for illumination and rotation invariant face detection is also presented. The originality of this work is derived from a variety of novel deep learning paradigms designed for the training of such architectures.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available