“Natural motion is Good motion” would be the idea that many researchers would agree with. However, compared to the vast and diverse studies to realize naturalness in agent or robot motion, the active discussion towards natural motion generation seems insufficient. Perhaps, researchers would give different answers to the question of what natural motion is.
Therefore, from this workshop, we would like to provide a forum for researchers who have worked or are interested in “natural motion generation” across various fields from industry and academia.
In this workshop, researchers from various backgrounds (i.e., graphics, virtual human, robotics, humanoids, legged robot, manipulator, human-robot interaction) will share and discuss attempts they have made for natural motion generation.
In addition, we would review and reinvestigate fundamental questions about “natural motion” as below:
- How can we define the “naturalness” in motion?
- Is Animal/Human-like motion all we need, or what else?
- What would be the benefit from “natural motion”?
- Does “naturalness” help in both task-oriented and task-agnostic motion generation?
- Which factor affects the “naturalness” of motion?
- Mechanical design of a robot, Human-likeness, Context-awareness of the motion, or what else?
We consider various topics of natural motion generation as below,
but not limited to:
- Task-oriented natural motion generation:
- Natural motion for object grasping or manipulation.
- Natural motion for locomotion.
- Task-agnostic natural motion generation:
- Natural motion for gesture or expression.
- Natural motion for interaction with humans.
- Bio-inspired/Human-inspired natural motion.
- Fundamental questions for natural motion
- What is it, Why do we need it, When do we need it, and How can we make and use it.
- Mechanical design for natural motion.
- Advanced algorithms for natural motion.
Hyemin Ahn, UNIST