PROJECT TITLE :

Robust Representation and Recognition of Facial Emotions Using Extreme Sparse Learning - 2015

ABSTRACT:

Recognition of natural emotions from human faces is an fascinating topic with a wide range of potential applications, like human-laptop interaction, automated tutoring systems, image and video retrieval, sensible environments, and driver warning systems. Traditionally, facial emotion recognition systems are evaluated on laboratory controlled information, that is not representative of the surroundings faced in real-world applications. To robustly recognize the facial emotions in real-world natural things, this paper proposes an approach called extreme sparse learning, which has the power to jointly learn a dictionary (set of basis) and a nonlinear classification model. The proposed approach combines the discriminative power of maximum learning machine with the reconstruction property of sparse illustration to enable accurate classification when presented with noisy signals and imperfect data recorded in natural settings. Likewise, this paper presents a replacement local spatio-temporal descriptor that's distinctive and create-invariant. The proposed framework is able to achieve the state-of-the-art recognition accuracy on both acted and spontaneous facial emotion databases.