aiDance: A Non-Invasive Approach in Designing AI-Based Feedback for Ballet Assessment and Learning

Abstract

Since its codified genesis in the 18th century, ballet training has largely been unchanged: it relies on tools that lack adequate support for both dancers and teachers. In particular, providing effective augmented feedback remains challenging as it can be limited, not always provided at the proper time, and highly subjective as it depends on the visual experience of an instructor. Designing a ballet assessment and learning tool with the aim of achieving a meaningful educational experience is an interdisciplinary challenge due to the fine motor movements and patterns of the art form. My work examines how we can effectively augment ballet learning in three phases using mixedmethod approaches. First, through my past professional experience as a ballet dancer, I explore how the design and in-lab evaluation of augmented visual and verbal feedback can improve the technical performance for novices and experts via remote learning. Second, I investigate the learning and teaching challenges that currently exist in traditional in-person training environments for dancers and teachers. Furthermore, I study the current technology use, reasons for non-use, and derive design requirements for future use. Lastly, I focus on how we can design aiDance, an AI-based feedback tool that attempts to represent an affordable and non-invasive approach that augments teachers’ abilities to facilitate assessment in the 21st century and pirouette towards the enhancement of learning. With this empirical work, I present insights that inform the HCI community at the intersection of dance and design in addressing the first steps towards the standardization of motor learning feedback.