Simple Baseline for Single Human Motion Forecasting
This work addresses the problem of computational efficiency in human motion forecasting for researchers, though it appears incremental as it builds on existing benchmarks with training tricks.
The paper tackles single human motion forecasting by establishing a simple baseline without visual or social information, achieving a large margin of improvement over existing methods on the SoMoF benchmark.
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.