Design and Evaluation of 3D Systems for Movement Transmission in Dance

Keywords :

• computer animation
• dance
• movement visualization
• ribbons
• HCI

In this thesis, I aim to design and evaluate 3D systems for movement transmission in dance. I explore dance transmission through two dimensions: transmission of specific aspects of a dance style to dancers and transmission of the animated dance from one virtual animated character to another.
When transmitting specific aspects of a dance style to dancers, I collaborated with a connoisseur dancer in Isadora Duncan’s style to create 3D systems in augmented reality. A first contribution is the co-design, technical implementation, and evaluation with dancers of an artifact that transmits the fluidity of movement. This system visualizes star-shaped ribbons joined at the solar plexus animated from motion-captured data of Isadora Duncan’s dances. The animation is performed using an optimization-based algorithm. The evaluation took place as a series of workshops. It showed great potential for the system to transmit the qualities of Isadora Duncan’s movement that a human-like or skeleton-based representation fails to transmit.
Based on the feedback of the first contribution, I co-designed with the same connoisseur dancer another system that supports the perception of spatiality in dance in augmented reality. This second contribution lies in the design and evaluation of the system. Along with the expert dancer, I took inspiration from the Sutton notation to create a system that visualizes animated keyframes of the dance and the trace of the solar plexus in space. I then ran a structured observation to evaluate the differences in the number of keyframes displayed simultaneously. My results show the potential of augmented reality for pedagogical purposes.
The last contribution of this thesis lies in the second aspect of dance transmission: transmission of the animated dance from one virtual animated character to another. I created a system that takes as an input motion-captured data of a dance and transfers it to another body with a different morphology. To achieve this, I first generate ribbons from the motion-captured data (animated rig or set of markers). Then, I match each ribbon with an articulated chain of the target rig. Lastly, I apply the animation contained in the ribbons onto the target rig using an optimization algorithm. 

President:

Laurence Nigay (Université Grenoble-Alpes)

Directors:

• Mélina Skouras (Inria Grenoble )
• Sarah Fdili Alaoui (Université Paris Saclay )
• Emmanuelle Crépeau (Université Grenoble-Alpes )

Raporteurs:

• Antonio Camurri (Università degli Studi di Genova )
• Michael Neff (University of California, Davis )

Examinators:

• Caroline Larboulette (Université de Bretagne Sud, IRISA )
• Katerina El Raheb (Université du Péloponnèse )