Articulated Body Motion Capture by Stochastic Search

By Jonathan A. J. Deutscher, Ian David Reid. International Journal of Computer Vision, 2005

481 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/38199c8b99acd210b3be834d911303056f593f24

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

Citations in the corpus

Benchmarking modern multiprocessors (2011, background): “The bodytrack computer vision application is an Intel RMS workload which tracks a 3D pose of a marker-less human body with multiple cameras through an image sequence [5, 23].”
Motion capture using joint skeleton tracking and surface estimation (2009, methodology): “Besides stochastic approaches [24] [12], global optimization techniques [9] [15] have been also proposed to overcome the limitations of local optimization.”
Combining multiple depth cameras and projectors for interactions on, above and between surfaces (2010, background): “Robust hand tracking and skeletal tracking can be quite difficult, and has been the subject of much research effort [7,30,31].”
VNect (2017, background): “monocular tracking of simple motions, but requires more expensive offline computation [de La Gorce et al. 2008]. Samplingbased methods avoid local minima [Balan et al. 2005; Bo and Sminchisescu 2010; Deutscher and Reid 2005; Gall et al. 2010]. However, real-time variants can not guarantee global con”
The TUM Kitchen Data Set of everyday manipulation activities for motion tracking and action recognition (2009, methodology): “Several systems have been presented [8, 10], yet most”
Markerless motion capture of interacting characters using multi-view image segmentation (2011, methodology): “Popular methods like [8, 13, 4] model a human’s motion by articulated bone hierarchies.”
A Nonrigid Kernel-Based Framework for 2D-3D Pose Estimation and 2D Image Segmentation (2011, unknown): “To overcome this constraint, nonrigid algorithms have appeared in the area of human pose estimation [17], [38], [1].”
HumanEva: Synchronized Video and Motion Capture Dataset for Evaluation of Articulated Human Motion (2006, background): “A variety of statistical [1, 2, 3, 6, 12, 36, 37, 38] as well as deterministic methods [21, 41, 35] have been developed for tracking people from single [1, 2, 7, 12, 15, 20, 21, 25, 26, 27, 29, 37] as well as multiple [3, 6, 9, 36] views.”
HumanEva: Synchronized Video and Motion Capture Dataset and Baseline Algorithm for Evaluation of Articulated Human Motion (2010, methodology): “The MSE between the predicted and observed silhouette values for these points is computed (Deutscher and Reid 2005): − logpf (yt |xt ) ∝ 1|{ξfxt }| ∑ j ( 1 − Mft ( ξ f xt (j ) ))2 .”
Vision-based human motion analysis: An overview (2007, background): “Deutscher and Reid [21] introduce a diﬀerent approach to guide the particles.”
Measure Locally, Reason Globally: Occlusion-sensitive Articulated Pose Estimation (2006, methodology): “These patches are connected in a kinematic tree [1, 2, 4, 11, 14, 18, 20] and the pose of the person is represented by a high-dimensional state vector that includes the position and orientation of the root limb in the global image coordinate frame and the parameters of each limb relative to its pare”
Dynamical binary latent variable models for 3D human pose tracking (2010, background): “Early dynamical models were formulated as smooth, linear Markov models [2] [3] [18], but such models do not capture nonlinear properties of human motion and were found insufficient for monocular 3D pose tracking.”
The Vitruvian manifold: Inferring dense correspondences for one-shot human pose estimation (2012, background): “However, it is almost invariably the case that high-capacity models have many local minima, meaning that finding a good minimum requires either expensive global search [13, 14], depends on a good initial estimate, or is applicable only to a limited range of poses.”
Estimating human shape and pose from a single image (2009, methodology): “There are several methods for representing body shape with varying levels of specificity: 1) non-parametric models such as visual hulls, point clouds and voxel representations (not considered further here); 2) part-based models using generic shape primitives such as cylinders or cones [9], superquad”