Long Term Motion Prediction Using Keyposes

TL;DR

This paper introduces a method for long-term human motion prediction by forecasting keyposes and interpolating intermediate poses, enabling more realistic and diverse predictions up to 5 seconds ahead.

Contribution

It proposes predicting keyposes instead of every pose for long-term forecasting, improving realism and diversity over extended time horizons.

Findings

Predicts realistic motions up to 5 seconds ahead

Generates multiple plausible future motions via probabilistic modeling

Outperforms state-of-the-art methods in realism and diversity

Abstract

Long term human motion prediction is essential in safety-critical applications such as human-robot interaction and autonomous driving. In this paper we show that to achieve long term forecasting, predicting human pose at every time instant is unnecessary. Instead, it is more effective to predict a few keyposes and approximate intermediate ones by interpolating the keyposes. We demonstrate that our approach enables us to predict realistic motions for up to 5 seconds in the future, which is far longer than the typical 1 second encountered in the literature. Furthermore, because we model future keyposes probabilistically, we can generate multiple plausible future motions by sampling at inference time. Over this extended time period, our predictions are more realistic, more diverse and better preserve the motion dynamics than those state-of-the-art methods yield.