Vision-based human motion analysis: An overview

By Ronald Poppe. Computer Vision and Image Understanding, 2007

960 citations recorded by the source corpus.

Markerless vision-based human motion analysis has the potential to provide an inexpensive, non-obtrusive solution for the estimation of body poses. The significant research effort in this domain has been motivated by the fact that many application areas, including surveillance, Human-Computer Interaction and automatic annotation, will benefit from a robust solution. In this paper, we discuss the characteristics of human motion analysis. We divide the analysis into a modeling and an estimation phase. Modeling is the construction of the likelihood function, estimation is concerned with finding the most likely pose given the likelihood surface. We discuss model-free approaches separately. This taxonomy allows us to highlight trends in the domain and to point out limitations of the current state of the art.

Subfields: Computer Vision and Pattern Recognition, Human-Computer Interaction

View Paper: https://www.semanticscholar.org/paper/ea481ceaf3ad8bef871a9efdddb27c345e0c3b4e

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

In Human-Computer Interaction

Citations in the corpus

Resolving 3D Human Pose Ambiguities With 3D Scene Constraints (2019, background): “The community has made significant progress on estimating human body pose and shape from images [18, 43, 53, 60].”
Pose-Sensitive Embedding by Nonlinear NCA Regression (2010, background): “Many techniques rely on sliding-window part detectors based on color and other features applied to controlled recording conditions ([10, 39, 4, 28] to name a few, we refer to [32] for a complete survey).”
Hierarchical Aligned Cluster Analysis for Temporal Clustering of Human Motion (2013, background): “S YSTEMS that can detect, recognize, and synthesize human motion are of interest in both research and industry due to the large number of potential applications in virtual reality, smart surveillance systems for advanced user interfaces, and motion analysis (see [1], [2], [3] for a review).”
Capturing the human figure through a wall (2015, background): “On the other hand, past markerless methods use cameras and infrared-based techniques – including Kinect, multi-view cameras, moving cameras, and time-of-ﬂight cameras – and require a direct line-of-sight from the sensor to the person’s body, e.g., [Shotton et al. 2013; Ganapathi et al. 2010; Hasler ”
A survey on vision-based human action recognition (2010, background): “The interest in the topic is motivated by the promise of many applications, both offline and online.”
Real time motion capture using a single time-of-flight camera (2010, methodology): “We describe a novel algorithm for propagating this noisy evidence about body part locations up the kinematic chain using the unscented transform.”
Interactive Markerless Articulated Hand Motion Tracking Using RGB and Depth Data (2013, methodology): “Several methods produce a 3D mesh and/or kinematic skeleton as their output [13, 17].”
Monocular Pedestrian Detection: Survey and Experiments (2009, background): “The authors of [21], [47], [57] cover methods for people detection, body pose estimation, and activity recognition .”
A survey of video datasets for human action and activity recognition (2013, background): “Contents lists available at SciVerse ScienceDirect Computer Vision and Image Understanding journal homepage: www.elsevier .com/ locate/cviu A survey of video datasets for human action and activity recognition q Jose M. Chaquet a,⇑, Enrique J. Carmona a, Antonio Fernández-Caballero b a Dpto. de Intel”
Pose-based human action recognition via sparse representation in dissimilarity space (2014, methodology): “In [25] Poppe provided taxonomy of this type of methods , classifying the approaches into model-based and model-free, depending on whether an a priori body model is employed or not.”
Joint Training of a Convolutional Network and a Graphical Model for Human Pose Estimation (2014, background): “We refer to [26] for a complete survey of this era.”
Bayesian Image Based 3D Pose Estimation (2016, background): “Human pose estimation (HPE) has been extensively studied during the years by considering videos, 2D images and depth data, [3–5].”
Monocular human pose estimation: A survey of deep learning-based methods (2020, background): “We refer to (Poppe, 2007; Gong et al., 2016) for a more detailed introduction.”
Free-Viewpoint Video of Human Actors Using Multiple Handheld Kinects (2013, background): “The last are markerless motion capture approaches that concentrate on capturing the skeleton motion of the human performance, e.g., [27], [28], [29], [30], [31].”