# Skeleton-Graph: Long-Term 3D Motion Prediction From 2D Observations Using Deep Spatio-Temporal Graph CNNs

Source: [Skeleton-Graph: Long-Term 3D Motion Prediction From 2D Observations Using Deep Spatio-Temporal Graph CNNs](<https://www.sota2.com/research/paper/68fa6b7a1e209131cea2e6f9>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-01-04T03:37:33.573Z

Authors: Abduallah Mohamed, Huancheng Chen, Zhangyang Wang, Christian Claudel

Published: 2021-09-21T15:33:40.000Z

## Abstract

Several applications such as autonomous driving, augmented reality and virtual reality require a precise prediction of the 3D human pose. Recently, a new problem was introduced in the field to predict the 3D human poses from observed 2D poses. We propose Skeleton-Graph, a deep spatio-temporal graph CNN model that predicts the future 3D skeleton poses in a single pass from the 2D ones. Unlike prior works, Skeleton-Graph focuses on modeling the interaction between the skeleton joints by exploiting their spatial configuration. This is being achieved by formulating the problem as a graph structure while learning a suitable graph adjacency kernel. By the design, Skeleton-Graph predicts the future 3D poses without divergence in the long-term, unlike prior works. We also introduce a new metric that measures the divergence of predictions in the long term. Our results show an FDE improvement of at least 27% and an ADE of 4% on both the GTA-IM and PROX datasets respectively in comparison with prior works. Also, we are 88% and 93% less divergence on the long-term motion prediction in comparison with prior works on both GTA-IM and PROX datasets. Code is available at https://github.com/abduallahmohamed/Skeleton-Graph.git

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [3D Pose Prediction on PROX](<https://www.sota2.com/research/sota/3d-pose-prediction-on-prox.md>): 3D Pose Prediction; PROX; 3D Pose Error (0.5s): 281.

- [3D Path Prediction on PROX](<https://www.sota2.com/research/sota/3d-path-prediction-on-prox.md>): 3D Path Prediction; PROX; Path Error (0.5s): 264.

- [3D Pose Prediction on GTA-IM (test)](<https://www.sota2.com/research/sota/3d-pose-prediction-on-gta-im-test.md>): 3D Pose Prediction; GTA-IM (test); 3D Pose Error (0.5s) (mm): 198.

- [3D Path Prediction on GTA-IM (test)](<https://www.sota2.com/research/sota/3d-path-prediction-on-gta-im-test.md>): 3D Path Prediction; GTA-IM (test); FDE: 208.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2109.10257>)

[PDF](<https://arxiv.org/pdf/2109.10257.pdf>)

[Implementation](<https://github.com/abduallahmohamed/Skeleton-Graph.git>)