MotionMap: Representing Multimodality in Human Pose Forecasting

TL;DR

MotionMap introduces a heatmap-based representation for multimodal human pose forecasting, enabling efficient sampling of multiple plausible futures, confidence estimation, and capturing rare modes, thus improving evaluation and controllability.

Contribution

The paper proposes MotionMap, a novel heatmap-based method that effectively models multimodality in human pose forecasting, addressing evaluation challenges and capturing rare, critical modes.

Findings

MotionMap accurately captures multiple forecast modes.

It provides confidence measures for different predictions.

The method demonstrates strong performance on Human3.6M and AMASS datasets.

Abstract

Human pose forecasting is inherently multimodal since multiple futures exist for an observed pose sequence. However, evaluating multimodality is challenging since the task is ill-posed. Therefore, we first propose an alternative paradigm to make the task well-posed. Next, while state-of-the-art methods predict multimodality, this requires oversampling a large volume of predictions. This raises key questions: (1) Can we capture multimodality by efficiently sampling a smaller number of predictions? (2) Subsequently, which of the predicted futures is more likely for an observed pose sequence? We address these questions with MotionMap, a simple yet effective heatmap based representation for multimodality. We extend heatmaps to represent a spatial distribution over the space of all possible motions, where different local maxima correspond to different forecasts for a given observation. MotionMap can capture a variable number of modes per observation and provide confidence measures for different modes. Further, MotionMap allows us to introduce the notion of uncertainty and controllability over the forecasted pose sequence. Finally, MotionMap captures rare modes that are non-trivial to evaluate yet critical for safety. We support our claims through multiple qualitative and quantitative experiments using popular 3D human pose datasets: Human3.6M and AMASS, highlighting the strengths and limitations of our proposed method. Project Page: https://vita-epfl.github.io/MotionMap