Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

Personalized Pose Forecasting

Priisalu, Maria LU ; Kronvall, Ted LU and Sminchisescu, Cristian LU (2023)
Abstract
Human pose forecasting is the task of predicting articulated human motion given past human motion. There exists a number of popular benchmarks that evaluate an array of different models performing human pose forecasting. These benchmarks do not reflect that a human interacting system, such as a delivery robot, observes and plans for the motion of the same individual over an extended period of time. Every individual has unique and distinct movement patterns. This is however not reflected in existing benchmarks that evaluate a model's ability to predict an average human's motion rather than a particular individual's. We reformulate the human motion forecasting problem and present a model-agnostic personalization method. Motion forecasting... (More)
Human pose forecasting is the task of predicting articulated human motion given past human motion. There exists a number of popular benchmarks that evaluate an array of different models performing human pose forecasting. These benchmarks do not reflect that a human interacting system, such as a delivery robot, observes and plans for the motion of the same individual over an extended period of time. Every individual has unique and distinct movement patterns. This is however not reflected in existing benchmarks that evaluate a model's ability to predict an average human's motion rather than a particular individual's. We reformulate the human motion forecasting problem and present a model-agnostic personalization method. Motion forecasting personalization can be performed efficiently online by utilizing a low-parametric time-series analysis model that personalizes neural network pose predictions. (Less)
Please use this url to cite or link to this publication:
author
; and
organization
publishing date
type
Working paper/Preprint
publication status
published
subject
keywords
Pose estimation
publisher
arXiv.org
DOI
10.48550/arXiv.2312.03528
language
English
LU publication?
yes
id
1231ca89-275d-45ac-85b2-91e50716d5d9
date added to LUP
2025-03-11 11:12:19
date last changed
2025-04-14 11:59:40
@misc{1231ca89-275d-45ac-85b2-91e50716d5d9,
  abstract     = {{Human pose forecasting is the task of predicting articulated human motion given past human motion. There exists a number of popular benchmarks that evaluate an array of different models performing human pose forecasting. These benchmarks do not reflect that a human interacting system, such as a delivery robot, observes and plans for the motion of the same individual over an extended period of time. Every individual has unique and distinct movement patterns. This is however not reflected in existing benchmarks that evaluate a model's ability to predict an average human's motion rather than a particular individual's. We reformulate the human motion forecasting problem and present a model-agnostic personalization method. Motion forecasting personalization can be performed efficiently online by utilizing a low-parametric time-series analysis model that personalizes neural network pose predictions.}},
  author       = {{Priisalu, Maria and Kronvall, Ted and Sminchisescu, Cristian}},
  keywords     = {{Pose estimation}},
  language     = {{eng}},
  month        = {{12}},
  note         = {{Preprint}},
  publisher    = {{arXiv.org}},
  title        = {{Personalized Pose Forecasting}},
  url          = {{http://dx.doi.org/10.48550/arXiv.2312.03528}},
  doi          = {{10.48550/arXiv.2312.03528}},
  year         = {{2023}},
}