A large, crowdsourced evaluation of gesture generation systems on common data: The GENEA Challenge 2020

TL;DR

The GENEA Challenge 2020 provided a large-scale, crowdsourced evaluation of gesture generation systems on a common dataset, enabling direct comparison and benchmarking of recent data-driven approaches in co-speech gesture synthesis.

Contribution

This paper introduces the GENEA Challenge 2020, establishing a standardized benchmark for evaluating and comparing gesture generation systems using a shared dataset and evaluation pipeline.

Findings

Systems showed varied performance, highlighting strengths and weaknesses.

Crowdsourced evaluation provided diverse user preferences.

Benchmarking revealed the current state and gaps in gesture generation methods.

Abstract

Co-speech gestures, gestures that accompany speech, play an important role in human communication. Automatic co-speech gesture generation is thus a key enabling technology for embodied conversational agents (ECAs), since humans expect ECAs to be capable of multi-modal communication. Research into gesture generation is rapidly gravitating towards data-driven methods. Unfortunately, individual research efforts in the field are difficult to compare: there are no established benchmarks, and each study tends to use its own dataset, motion visualisation, and evaluation methodology. To address this situation, we launched the GENEA Challenge, a gesture-generation challenge wherein participating teams built automatic gesture-generation systems on a common dataset, and the resulting systems were evaluated in parallel in a large, crowdsourced user study using the same motion-rendering pipeline. Since differences in evaluation outcomes between systems now are solely attributable to differences between the motion-generation methods, this enables benchmarking recent approaches against one another in order to get a better impression of the state of the art in the field. This paper reports on the purpose, design, results, and implications of our challenge.