AnyPose: Anytime 3D Human Pose Forecasting via Neural Ordinary Differential Equations

09/09/2023

∙

Anytime 3D human pose forecasting is crucial to synchronous real-world human-machine interaction, where the term “anytime" corresponds to predicting human pose at any real-valued time step. However, to the best of our knowledge, all the existing methods in human pose forecasting perform predictions at preset, discrete time intervals. Therefore, we introduce AnyPose, a lightweight continuous-time neural architecture that models human behavior dynamics with neural ordinary differential equations. We validate our framework on the Human3.6M, AMASS, and 3DPW dataset and conduct a series of comprehensive analyses towards comparison with existing methods and the intersection of human pose and neural ordinary differential equations. Our results demonstrate that AnyPose exhibits high-performance accuracy in predicting future poses and takes significantly lower computational time than traditional methods in solving anytime prediction tasks.

READ FULL TEXT

AnyPose: Anytime 3D Human Pose Forecasting via Neural Ordinary Differential Equations

Forecasting Human Dynamics from Static Images

Action-Agnostic Human Pose Forecasting

Imitation Learning for Human Pose Prediction

Human Pose Forecasting via Deep Markov Models

Simple Baseline for Single Human Motion Forecasting

Toward Reliable Human Pose Forecasting with Uncertainty

ProtoRes: Proto-Residual Architecture for Deep Modeling of Human Pose

AnyPose: Anytime 3D Human Pose Forecasting via Neural Ordinary Differential Equations

Related Research

Forecasting Human Dynamics from Static Images

Action-Agnostic Human Pose Forecasting

Imitation Learning for Human Pose Prediction

Human Pose Forecasting via Deep Markov Models

Simple Baseline for Single Human Motion Forecasting

Toward Reliable Human Pose Forecasting with Uncertainty

ProtoRes: Proto-Residual Architecture for Deep Modeling of Human Pose