Human Pose Tracking Using Multi-level Structured Models
Lecture notes in computer sciencePublished 1 January 2006
Mun Wai Lee, Ram Nevatia
Citations23
SJR quartileQ2
SJR score0.35
SNIP0.55
Generate an AI Snapshot to get a quick, structured summary of this paper.
Study Snapshot
ObjectiveStudy objective
MethodsResearch methodology
PopulationPopulation studied
Sample sizeSample sizes
OutcomesStudy outcomes here
ResultsStudy results comes here
LimitationsResearch study limitations comes here
A concise AI-generated summary of the paper will appear here once you click Generate AI Snapshot.
TL;DR
Experimental results on realistic indoor video sequences show that the proposed three-stage approach with a multi-level state representation that enables a hierarchical estimation of 3D body poses is able to track multiple persons during complex movement such as turning movement with inter-occlusion.
Abstract
Tracking body poses of multiple persons in monocular video is a challenging problem due to the high dimensionality of the state space and issues such as inter-occlusion of the persons' bodies. We proposed a three-stage approach with a multi-level state representation that enables a hierarchical estimation of 3D body poses. At the first stage, humans are tracked as blobs. In the second stage, parts such as face, shoulders and limbs are estimated and estimates are combined by grid-based belief propagation to infer 2D joint positions. The derived belief maps are used as proposal functions in the third stage to infer the 3D pose using data-driven Markov chain Monte Carlo. Experimental results on realistic indoor video sequences show that the method is able to track multiple persons during complex movement such as turning movement with inter-occlusion.
Keywords
Computer ScienceEngineering
Markov Chain Monte Carlo in Practice
7,730 Citations1995Walter R. Gilks, Sylvia Richardson +1 more
Adaptive background mixture models for real-time tracking
6,930 Citations2003Chris Stauffer, W. Eric L. Grimson
This paper discusses modeling each pixel as a mixture of Gaussians and using an on-line approximation to update the model, resulting in a stable, real-time outdoor tracker which reliably deals with lighting changes, repetitive motions from clutter, and long-term scene changes.
Fast pose estimation with parameter-sensitive hashing
764 Citations2003Shakhnarovich, Viola +1 more
A new algorithm is introduced that learns a set of hashing functions that efficiently index examples relevant to a particular estimation task, and can rapidly and accurately estimate the articulated pose of human figures from a large database of example images.
IEEE Transactions on Pattern Analysis and Machine IntelligenceImage segmentation by data-driven markov chain monte carlo
567 Citations2002Zhuowen Tu, Song-Chun Zhu
The DDMCMC paradigm provides a unifying framework in which the role of many existing segmentation algorithms are revealed as either realizing Markov chain dynamics or computing importance proposal probabilities and generalizes these segmentation methods in a principled way.
Recovering human body con .gurations:combining segmentation and recognition
504 Citations2004Giulio Mori, Xiaofeng Ren +2 more
This work uses segmentation to build limb and torso detectors, the outputs of which are assembled into human figures, and presents quantitative results on torso localization, in addition to shortlisted full body configurations.
Nonparametric belief propagation
422 Citations2003Erik B. Sudderth, Alexander Ihler +2 more
Tracking loose-limbed people
391 Citations2004Leonid Sigal, Shalini Bhatia +3 more
The problem of 3D human tracking as one of inference in a graphical model that is a collection of loosely-connected limbs and non-parametric belief propagation using a variation of particle filtering that can be applied over a general loopy graph is posed.
3D human pose from silhouettes by relevance vector regression
389 Citations2004Ankur Agarwal, Bill Triggs
This work describes a learning based method for recovering 3D human body pose by direct nonlinear regression against shape descriptor vectors extracted automatically from image silhouettes, and results are a factor of 3 better than the current state of the art for the much simpler upper body problem.
Tracking multiple humans in crowded environment
338 Citations2004Tao Zhao, R. Nevatia
A method that can track humans in crowded environments, with significant and persistent occlusion by making use of human shape models in addition to camera models, the assumption that humans walk on a plane and acquired appearance models is presented.
Kinematic jump processes for monocular 3D human tracking
270 Citations2003Cristian Sminchisescu, Bill Triggs
This work uses simple kinematic reasoning to enumerate the tree of possible forwards/backwards flips, thus greatly speeding the search within each linked group of minima in the model-image matching cost function.
Finding and tracking people from the bottom up
269 Citations2003Deva Ramanan, David Forsyth
A tracker that can track moving people in long sequences without manual initialization is described and it is shown the tracking algorithm can be interpreted as a loopy inference procedure on an underlying Bayes net.
PAMPAS: real-valued graphical models for computer vision
191 Citations2003Michael Isard
This paper combines belief propagation with ideas from particle filtering; the resulting algorithm performs inference on graphs containing both cycles and continuous-valued latent variables with general conditional probability distributions, which has wide applicability in the computer vision domain.
Proposal maps driven MCMC for estimating human body pose in static images
147 Citations2004Mun Wai Lee, I. Cohen
An image generative approach is used by modeling the human kinematics, the shape and the clothing probabilistically, used for deriving a good likelihood measure to evaluate samples in the solution, space and introduces the use of 'proposal maps', which is an efficient way of consolidating the evidence and generating 3D pose candidates during the MCMC search.
Automatic partitioning of high dimensional search spaces associated with articulated body motion capture
141 Citations2005J. Deutscher, Andrew J. Davison +1 more
This work develops a hierarchical search strategy which automatically partitions the search space without any explicit representation of the partitions and introduces a crossover operator (similar to that found in genetic algorithms) which improves the ability of the tracker to search different partitions in parallel.
Tracking articulated body by dynamic Markov network
117 Citations2003Ying Wu, Gang Hua +1 more
A new method for visual tracking of articulated objects that analyzes subparts locally while reinforcing the structural constraints at the mean time is presented and Probabilistic variational analysis of the model reveals a mean field approximation to the posterior densities of each subparts given visual evidence.
Learning to Estimate Human Pose with Data Driven Belief Propagation
109 Citations2005Gang Hua, Ming–Hsuan Yang +1 more
A data driven belief propagation Monte Carlo algorithm, utilizing importance sampling functions built from bottom-up visual cues, is proposed for efficient probabilistic inference for 2D human pose estimation from single images.
Distributed Occlusion Reasoning for Tracking with Nonparametric Belief Propagation
100 Citations2004Erik B. Sudderth, Michael Mandel +2 more
A three-dimensional geometric hand model suitable for visual tracking applications is described and it is shown that local structure may be recovered by introducing binary hidden variables describing the occlusion state of each pixel.
Integrating bottom-up/top-down for object recognition by data driven Markov chain Monte Carlo
78 Citations2002Song‐Chun Zhu, Rong Zhang +1 more
This article presents a mathematical paradigm called Data Driven Markov Chain Monte Carlo (DDMCMC) for object recognition that realizes traditional "hypothesis-and-test" methods through well-balanced Markov chain Monte Carlo dynamics, thus it achieves robust and globally optimal solutions.
The NIST Meeting Room Pilot Corpus
75 Citations2004John S. Garofolo, Christophe Laprun +3 more
A multi-media corpus of audio and video from meetings collected at NIST using a variety of microphones and video cameras is being used in the development and evaluation of speech technologies and by the video extraction community under the auspices of the ARDA Video Analysis and Content Exploitation (VACE) program.
Representation and matching of articulated shapes
52 Citations2004Jiayong Zhang, Robert T. Collins +1 more
This work considers the problem of localizing the articulated and deformable shape of a walking person in a single view by representing the non-rigid 2D body contour by a Bayesian graphical model whose nodes correspond to point positions along the contour.
Lecture notes in computer scienceHuman Pose Estimation Using Learnt Probabilistic Region Similarities and Partial Configurations
48 Citations2004Timothy J. Roberts, S.J. McKenna +1 more
A model of human appearance is presented for efficient pose estimation from real-world images using a high-level model which defines a space of configurations which can be associated with image measurements and thus scored.
Dynamic Human Pose Estimation using Markov Chain Monte Carlo Approach
30 Citations2005Mun Wai Lee, Ramakant Nevatia
A method based on data-driven Markov chain Monte Carlo (DD-MCMC) that uses bottom-up techniques to generate state proposals for pose estimation and initialization including automatic pose initialization and re-initialization after tracking failures caused by partial occlusion or unreliable observations is proposed.