A Human-Computer Duet System for Music Performance

TL;DR

This paper presents a virtual violinist system capable of collaborating with human pianists in real-time chamber music performances, integrating music tracking, pose estimation, and movement generation for seamless human-virtual duet performances.

Contribution

It introduces a novel virtual musician system that autonomously generates performance behaviors based solely on audio input, enabling natural human-virtual collaboration.

Findings

System successfully performed in public concerts

Achieved low-cost and scalable co-performance

Validated with objective quality assessments

Abstract

Virtual musicians have become a remarkable phenomenon in the contemporary multimedia arts. However, most of the virtual musicians nowadays have not been endowed with abilities to create their own behaviors, or to perform music with human musicians. In this paper, we firstly create a virtual violinist, who can collaborate with a human pianist to perform chamber music automatically without any intervention. The system incorporates the techniques from various fields, including real-time music tracking, pose estimation, and body movement generation. In our system, the virtual musician's behavior is generated based on the given music audio alone, and such a system results in a low-cost, efficient and scalable way to produce human and virtual musicians' co-performance. The proposed system has been validated in public concerts. Objective quality assessment approaches and possible ways to systematically improve the system are also discussed.