Submission Data for SoMoFormer




Human pose forecasting is a challenging problem involving complex human body motion and posture dynamics. In cases that there are multiple people in the environment, one's motion may also be influenced by the motion and dynamic movements of others. Although there are several previous works targeting the problem of multi-person dynamic pose forecasting, they often model the entire pose sequence as time series (ignoring the underlying relationship between joints) or only output the future pose sequence of one person at a time. In this paper, we present a new method, called Social Motion Transformer (SoMoFormer), for multi-person 3D pose forecasting. Our transformer architecture uniquely models human motion input as a joint sequence rather than a time sequence, allowing us to perform attention over joints while predicting an entire future motion sequence for each joint in parallel. We show that with this problem reformulation, SoMoFormer naturally extends to multi-person scenes by using the joints of all people in a scene as input queries. Using learned embeddings to denote the type of joint, person identity, and global position, our model learns the relationships between joints and between people, attending more strongly to joints from the same or nearby people. SoMoFormer outperforms state-of-the-art methods for long-term motion prediction on the SoMoF benchmark as well as the CMU-Mocap and MuPoTS-3D datasets. Code will be made available at

Publication title

SoMoFormer: Multi-Person Pose Forecasting with Transformers

Publication authors

Edward Vendrow, Satyajit Kumar, Ehsan Adeli, Hamid Rezatofighi

Publication venue and year


Publication URL




Source code URL

Creation date

2022-03-07 23:00:50

Metric 80/100 ms 160/240 ms 320/500 ms 400/640 ms 560/900 ms
PoseTrack VAM None None None None None
PoseTrack VIM None None None None None
3DPW VIM 9.096478634478709 21.309155299527312 47.50906156336407 61.566951475661824 91.89581997275748