Pedestrian Motion Trajectory Prediction With Stereo-Based 3D Deep Pose Estimation and Trajectory Learning

Zhong, Jianqi; Sun, Hao; Cao, Wenming; He, Zhihai

doi:10.1109/access.2020.2969994

Cited by 28 publications

(17 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Naturally, sports analytics is not the only domain where movement models are prominently deployed. Besides the already mentioned areas, other applications include vehicle trajectory analysis (Besse et al 2018), pedestrian trajectory estimation from videos (Zhong et al 2020), and real-time robot trajectory prediction (Gomez-Gonzalez et al 2020).…”

Section: Related Workmentioning

confidence: 99%

Contextual movement models based on normalizing flows

et al. 2021

View full text Add to dashboard Cite

Movement models predict positions of players (or objects in general) over time and are thus key to analyzing spatiotemporal data as it is often used in sports analytics. Existing movement models are either designed from physical principles or are entirely data-driven. However, the former suffers from oversimplifications to achieve feasible and interpretable models, while the latter relies on computationally costly, from a current point of view, nonparametric density estimations and require maintaining multiple estimators, each responsible for different types of movements (e.g., such as different velocities). In this paper, we propose a unified contextual probabilistic movement model based on normalizing flows. Our approach learns the desired densities by directly optimizing the likelihood and maintains only a single contextual model that can be conditioned on auxiliary variables. Training is simultaneously performed on all observed types of movements, resulting in an effective and efficient movement model. We empirically evaluate our approach on spatiotemporal data from professional soccer. Our findings show that our approach outperforms the state of the art while being orders of magnitude more efficient with respect to computation time and memory requirements.

show abstract

Section: Related Workmentioning

confidence: 99%

Contextual movement models based on normalizing flows

et al. 2021

View full text Add to dashboard Cite

show abstract

“…After obtained the transformation parameters, we can construct the transformation matrix T r according to Eqn. (1). Then, the transformed human pose can be calculated as following:…”

Section: A Viewpoint Transformationmentioning

confidence: 99%

“…Since there is no need of any specialized devices, like depth sensor, its possible application range is much wider than pose estimation from multi-views or RGB-D images. Due to its huge potential application in human motion prediction, action analysis and intelligent video surveillance [1]- [3], 3D human pose estimation from a monocular image has attracted more and more attention in recent years.…”

Section: Introductionmentioning

confidence: 99%

An Adaptive Viewpoint Transformation Network for 3D Human Pose Estimation

et al. 2020

View full text Add to dashboard Cite

Human pose estimation from a monocular image has attracted lots of interest due to its huge potential application in many areas. The performance of 2D human pose estimation has been improved a lot with the emergence of deep convolutional neural network. In contrast, the recovery of 3D human pose from an 2D pose is still a challenging problem. Currently, most of the methods try to learn a universal map, which can be applied for all human poses in any viewpoints. However, due to the large variety of human poses and camera viewpoints, it is very difficult to learn a such universal mapping from current datasets for 3D pose estimation. Instead of learning a universal map, we propose to learn an adaptive viewpoint transformation module, which transforms the 2D human pose to a more suitable viewpoint for recovering the 3D human pose. Specifically, our transformation module takes a 2D pose as input and predicts the transformation parameters. Rather than some hand-crafted criteria, this module is directly learned from the datasets and depends on the input 2D pose in testing phrase. Then the 3D pose is recovered from this transformed 2D pose. Since the difficulty of 3D pose recovery becomes smaller, we can obtain more accurate estimation results. Experiments on Human3.6M and MPII datasets show that the proposed adaptive viewpoint transformation can improve the performance of 3D human pose estimation. INDEX TERMS 3D human pose estimation, adaptive viewpoint transformation, deep convolutional neural network

show abstract

“…Ground interval contour estimation problem (henceforth, will be refered as ground contour estimation), can be viewed from another perspective, namely threedimensional (3D) trajectory estimation of a system. 3D trajectory estimation is one branch of computer science that is able to produce a collection of 3D point coordinates in space that are able to represent trace or movement of the system while the system is working [3,10].…”

Section: A Backgroundmentioning

confidence: 99%

Performance Evaluation of Direction Cosine Matrix and Madgwick’s as 3D Orientation Estimation Algorithm for Estimating Ground Contour

Wibowo

Purwanto

Christie

et al. 2020

Preprint

View full text Add to dashboard Cite

Abstract Contour interval is elevation difference between two contour lines that are adjacent and parallel. Ground Contour interval is an important information that must be provided in a construction. Ground contour interval estimation problem could be viewed as a 3D trajectory estimation problem. One framework, called dead reckoning, is often used for motion (position and orientation) tracking of a mobile systems over time. Dead reckoning requires a precise 3D orientation estimation algorithm. Two competing algorithms are Direction Cosine Matrix and Madgwick’s. To ensure the 3D trajectory estimation gives an accurate result, these two algorithms must be carefully and thoroughly evaluated. This research is part of a complete research on the development of a ground contour estimation and only focuses on comparison between two 3D orientation estimation algorithm.

show abstract

Pedestrian Motion Trajectory Prediction With Stereo-Based 3D Deep Pose Estimation and Trajectory Learning

Cited by 28 publications

References 22 publications

Contextual movement models based on normalizing flows

Contextual movement models based on normalizing flows

An Adaptive Viewpoint Transformation Network for 3D Human Pose Estimation

Performance Evaluation of Direction Cosine Matrix and Madgwick’s as 3D Orientation Estimation Algorithm for Estimating Ground Contour

Contact Info

Product

Resources

About