[robotics-worldwide] [journals]

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view

[robotics-worldwide] [journals]

Agnese Augello

About this Research Topic
Body language in social robots is essential to convey naturalness and ease
interaction; in short, to make them trustworthy. Human-like robots are more
appropriate to imitate human gesticulation behavior, especially when it
alludes to talking gesture generation.

Gesticulating while talking is innate in humans, and is more common in
certain cultures. Speech is often accompanied by moving the head, nodding,
and orienting the gaze. Particularly involved is the motion of the arms and
the hands.

While some gestures, such as beat ones, are used without particular meaning,
others can be used to convey different information, emphasize what people
are saying, point out something, and to communicate emotions. The gestures
can be unintended or intentional, with the latter often revealing aspects
related to personality and mood. Emotions cause a series of changes that can
also be noticeable in postures and gestures.

We naturally produce these rich gestures spontaneously without effort,
however, to obtain an automatic generation of gestures in robots, there are
several issues that must be tackled.

This Research Topic aims to gain a better understanding of emotional
gesturing in robotics by exploring how natural gesticulation is generated in
social robots and the types of training and input data required to generate
a variety of gestures. We also aim to identify which features of emotional
gestures should be emphasised in order to accurately represent the speaker's
emotion and/or tone.

Aside from co-speech gestures, body motion is also present in humans even
when listening or thinking, therefore, we also seek to assess whether robots
should also replicate these gestures to improve social experiences.

It is also important to develop methods for robustly and objectively
evaluating these kinds of behaviors. The produced gesturing motion should be
perceived as natural and acceptable by the general audience.
Notwithstanding, the lack of proper quantitative measures makes it
impossible to compare multiple systems. Hence, we also consider any
proposals for methods of objectively evaluating gesturing systems.

This Research Topic explores themes including, but not limited to:

• Human-like gesticulation in social robots
• Gestures generation
• Co-speech gesturing systems
• Multimodal expression of emotions
• Affective models for gesture generation
• Methodologies for evaluating robotic gestures
• Quantitative Measures for the evaluation of not verbal human-robot
• Context-based gesticulation
• Gestures classification
• Creation of dataset for multimodal human-robot interaction
• Real-time gesture Imitation
• Gesture learning
• Non-verbal interaction
• Computational creativity for gesture generation
• Social models for multimodal human-robot interaction

Keywords: Human-like gesticulation, Social Robots, Robot Gesticulation,
Human-Robot Interaction, Humanoid Robots

Sent from: https://urldefense.com/v3/__http://robotics-worldwide.1046236.n5.nabble.com/__;!!LIr3w8kk_Xxm!_Ci8uS4hqonS9QXAFy57H2lPbaslXECh_pkVndcRJRpJMPEJVw6QZan7fXt_Lmu_SNlC2ber$ 
robotics-worldwide mailing list
[hidden email]