A Spatio-temporal Transformer for 3D Human Motion Prediction


Loading...

Date

2021

Publication Type

Conference Paper

ETH Bibliography

yes

Citations

Altmetric

Data

Abstract

We propose a novel Transformer-based architecture for the task of generative modelling of 3D human motion. Previous work commonly relies on RNN-based models considering shorter forecast horizons reaching a stationary and often implausible state quickly. Recent studies show that implicit temporal representations in the frequency domain are also effective in making predictions for a predetermined horizon. Our focus lies on learning spatio-temporal representations autoregressively and hence generation of plausible future developments over both short and long term. The proposed model learns high dimensional embeddings for skeletal joints and how to compose a temporally coherent pose via a decoupled temporal and spatial selfattention mechanism. Our dual attention concept allows the model to access current and past information directly and to capture both the structural and the temporal dependencies explicitly. We show empirically that this effectively learns the underlying motion dynamics and reduces error accumulation over time observed in auto-regressive models. Our model is able to make accurate short-term predictions and generate plausible motion sequences over long horizons.

Publication status

published

Editor

Book title

2021 International Conference on 3D Vision (3DV)

Journal / series

Volume

Pages / Article No.

565 - 574

Publisher

IEEE

Event

9th International Conference on 3D Vision (3DV 2021)

Edition / version

Methods

Software

Geographic location

Date collected

Date created

Subject

Organisational unit

03979 - Hilliges, Otmar (ehemalig) / Hilliges, Otmar (former) check_circle

Notes

Conference lecture held on December 2, 2021

Funding

717054 - Optimization-based End-User Design of Interactive Technologies (EC)

Related publications and datasets