Simple Baseline for Single Human Motion Forecasting

TL;DR

This paper introduces a simple, resource-efficient baseline for single human motion forecasting that outperforms existing methods on the SoMoF benchmark without relying on visual or social cues.

Contribution

The authors propose a novel baseline method for human motion forecasting that avoids complex visual and social data, using effective training tricks to achieve superior performance.

Findings

Outperforms existing methods on SoMoF benchmark

Does not require visual or social information

Utilizes effective training tricks

Abstract

Global human motion forecasting is important in many fields, which is the combination of global human trajectory prediction and local human pose prediction. Visual and social information are often used to boost model performance, however, they may consume too much computational resource. In this paper, we establish a simple but effective baseline for single human motion forecasting without visual and social information, equipped with useful training tricks. Our method "futuremotion_ICCV21" outperforms existing methods by a large margin on SoMoF benchmark. We hope our work provide new ideas for future research.