TrajectoryNet: a new spatio-temporal feature learning network for human motion prediction

TL;DR

TrajectoryNet is a novel 2D CNN-based network that models complex spatio-temporal features to accurately predict human motion trajectories, outperforming existing methods on multiple benchmarks.

Contribution

The paper introduces TrajectoryNet, which effectively captures coupled spatio-temporal, local-global spatial, and global temporal co-occurrence features for improved human motion prediction.

Findings

Achieves state-of-the-art results on Human3.6M, G3D, and FNTU datasets.

Effectively models motion dynamics with coupled features.

Demonstrates superior performance compared to existing methods.

Abstract

Human motion prediction is an increasingly interesting topic in computer vision and robotics. In this paper, we propose a new 2D CNN based network, TrajectoryNet, to predict future poses in the trajectory space. Compared with most existing methods, our model focuses on modeling the motion dynamics with coupled spatio-temporal features, local-global spatial features and global temporal co-occurrence features of the previous pose sequence. Specifically, the coupled spatio-temporal features describe the spatial and temporal structure information hidden in the natural human motion sequence, which can be mined by covering the space and time dimensions of the input pose sequence with the convolutional filters. The local-global spatial features that encode different correlations of different joints of the human body (e.g. strong correlations between joints of one limb, weak correlations between joints of different limbs) are captured hierarchically by enlarging the receptive field layer by layer and residual connections from the lower layers to the deeper layers in our proposed convolutional network. And the global temporal co-occurrence features represent the co-occurrence relationship that different subsequences in a complex motion sequence are appeared simultaneously, which can be obtained automatically with our proposed TrajectoryNet by reorganizing the temporal information as the depth dimension of the input tensor. Finally, future poses are approximated based on the captured motion dynamics features. Extensive experiments show that our method achieves state-of-the-art performance on three challenging benchmarks (e.g. Human3.6M, G3D, and FNTU), which demonstrates the effectiveness of our proposed method. The code will be available if the paper is accepted.