Multiscale Residual Learning of Graph Convolutional Sequence Chunks for Human Motion Prediction

TL;DR

This paper introduces ResChunk, a novel end-to-end graph-based neural network that dynamically models spatio-temporal dependencies for human motion prediction, outperforming existing methods on benchmark datasets.

Contribution

ResChunk dynamically explores body component relationships and learns residuals between sequence chunks, addressing fixed scale limitations and mode collapse in human motion prediction.

Findings

Outperforms previous methods on CMU Mocap and Human3.6M datasets.

Effectively models dynamic spatio-temporal features.

Sets a new state-of-the-art in human motion prediction.

Abstract

A new method is proposed for human motion prediction by learning temporal and spatial dependencies. Recently, multiscale graphs have been developed to model the human body at higher abstraction levels, resulting in more stable motion prediction. Current methods however predetermine scale levels and combine spatially proximal joints to generate coarser scales based on human priors, even though movement patterns in different motion sequences vary and do not fully comply with a fixed graph of spatially connected joints. Another problem with graph convolutional methods is mode collapse, in which predicted poses converge around a mean pose with no discernible movements, particularly in long-term predictions. To tackle these issues, we propose ResChunk, an end-to-end network which explores dynamically correlated body components based on the pairwise relationships between all joints in individual sequences. ResChunk is trained to learn the residuals between target sequence chunks in an autoregressive manner to enforce the temporal connectivities between consecutive chunks. It is hence a sequence-to-sequence prediction network which considers dynamic spatio-temporal features of sequences at multiple levels. Our experiments on two challenging benchmark datasets, CMU Mocap and Human3.6M, demonstrate that our proposed method is able to effectively model the sequence information for motion prediction and outperform other techniques to set a new state-of-the-art. Our code is available at https://github.com/MohsenZand/ResChunk.