Invariant representation of facial expressions for blended expression recognition on unknown subjects

• Published in: Computer vision and image understanding Vol. 117; no. 11; pp. 1598 - 1609
• Main Authors: Soladié, Catherine, Stoiber, Nicolas, Séguier, Renaud
• Format: Journal Article
• Language: English
• Published: Amsterdam Elsevier Inc 01.11.2013; Elsevier
• Subjects: Applied sciences; Artificial intelligence; Blended; Computer Science; Computer science; control theory; systems; Delaunay tessellation; Engineering Sciences; Exact sciences and technology; Face recognition; Facial; Facial expression; Invariant representation; Invariants; Learning and adaptive systems; Manifold of expressions; Pattern recognition. Digital image processing. Computational geometry; Piecewise affine warping; Recognition; Representations; Signal and Image processing; Surface layer; Texture; Transfer learning; Piecewise affine warping; Delaunay tessellation; Manifold of expressions; Invariant representation; Transfer learning; Facial expression; Invariant; Data type; Computer vision; Dimensionality; Warping; Delaunay triangulation; Emotion emotionality; Knowledge management; Inductive learning; Texture; Knowledge transfer; Model matching; Image matching; Morphology; Facies; Tiling; Piecewise linearization
• ISSN: 1077-3142; 1090-235X
• DOI: 10.1016/j.cviu.2013.07.005

•The organization of facial expressions is person-independent.•One expression can be defined by its relative position to 8 other expressions.•We synthesized the 8 expressions from plausible distortions.•The representation is robust to the type of data (shape and/or texture information).•As expression is accurately described in a 4 dimensional expression space. Facial expressions analysis plays an important part in emotion detection. However, having an automatic and non-intrusive system to detect blended facial expression is still a challenging problem, especially when the subject is unknown to the system. Here, we propose a method that adapts to the morphology of the subject and that is based on a new invariant representation of facial expressions. In our system, one expression is defined by its relative position to 8 other expressions. As the mode of representation is relative, we show that the resulting expression space is person-independent. The 8 expressions are synthesized for each unknown subject from plausible distortions. Recognition tasks are performed in this space with a basic algorithm. The experiments have been performed on 22 different blended expressions and on either known or unknown subjects. The recognition results on known subjects demonstrate that the representation is robust to the type of data (shape and/or texture information) and to the dimensionality of the expression space. The recognition results on 22 expressions of unknown subjects show that a dimensionality of the expression space of 4 is enough to outperform traditional methods based on active appearance models and accurately describe an expression.