Research Topics
Facial Expression Recognition in Huma-robot Interaction
As human beings, we ought to express our ideas and emotions using several channels; texture, gait gesture, body language or facial expressions. a study made by Ferris reported that the facial expression of the speaker contributes 55% to the effect of the spoken message, while the verbal part and the vocal part contribute only 7% and 38% respectively. Facial expression was proven to be the most efficient and most powerful message transmitter of non-verbal communication form that shape a universal language and help individuals to recognize the emotional state of each other.
In today’s technology driven world, it is essential for human beings to adapt rapidly the fact that machines and robots are an integrated part of all aspects of life for the use by or with humans, at houses as a voice assistants, cleaner robots, security robots, in the cars as personalized driving systems that controls the vehicle according to our driving style…etc. The challenge to rise in this situation is the ability to fulfill some conditions to guarantee a natural, smooth and yet robust communication between machines and humans so it will meet the expectations to which those machines are designed for.
Social robots are one popular type of robots that aims to interact with human being in a direct way. and for that matter the multi-disciplinary field of research, Human robot interaction, invested in deeply analyzing the human-robot interaction with different age phases and conditions, children, adults, elderly, autistic. this lead to a major challenge and a question of how robots can have the human-like behavior in order to have a more effective communication, of course Human robot interaction may take several forms in terms of cooperation, duration and role.
In order to give the machines the ability to detect, recognize, interpret, process, and generate human like to humans, a research area has emerged, Affective computing. Affective computing is an interdisciplinary research study that knew an extensive attention recently, it is associated with computer science, cognitive science, neuroscience, sociology, psychology.
Another major factor that played as a link in human robot interaction and also made an evolutionary transformation in communication between robot-robot or human-robots is the dynamics of info-communication technology. and that was fulfilled by the emerge of both cloud computing and service robots in a way that this paradigm solved the problem of the limitation of hardware and speed of the single agent using cloud resources.
In this context, several questions arise about:
- How to improve the facial emotion recognition systems and get higher accuracy? How to judge those results?
- How to guarantee the long term communication between humans and robots, in terms of smoothness and natural?
- How to guarantee the effectiveness and speed of the facial emotion recognition and the rapidity of human robot interaction?
Motivation
In Computer vision, facial emotion recognition is classification of facial features in one of the six
basic universal emotions : happiness, sadness, fear, disgust, surprise and anger, as introduced by Ekman.
This classification allow machine to understand, simulate and react to human emotions in smooth way and rapid, especially with the rise of 5G that will guarantee the rapidity.
A motivation of the cloud based emotion recognition in human robot interaction has a variety of disciplines that will benefit from this phenomenon:
- Health care: many investments made in applying facial emotion
recognition in health care systems, to recognize and be aware about the emotional state that patients exhibit during their rehabilitation for a better assessment with treatment process by providing more attention to patients who need it. - Education: facial emotion recognition is used in education to have a better understanding of the adaption of learners to the study material, and based on the analysis, an adjustment of the teaching methodology is made.
- User feedback: monitoring the user’s and customers expressions while watching a movie, playing games, or do shopping can be critical for the industry to fundamentally understand the needs of users and customers and get feedback about their services or products for bigger profit and better marketing.
- Security and safety: applications in surveillance were designed based on facial emotion recognition in order to detect suspicious people.
Objectives
- Find a link between three research areas: Facial expression recognition, human robot interaction, cloud computing and cloud robotics.
- Answer the questions asked in the problematic section.
- Provide an extensive review about facial expression recognition, human robot interaction, cloud computing and cloud robotics.
- Give a discussion about approaches, models, algorithms used for facial emotion recognition.
- Propose scientific and technological goals based on the discussion, in order to make contributions in the conjunction of the three research fields.