Human Pose Estimation from Polluted Silhouettes Using Sub-manifold Voting Strategy
Lecture notes in computer sciencePublished 1 January 2006
Chunfeng Shen, Xueyin Lin, Yuanchun Shi
Citations1
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
A framework of human pose estimation from polluted silhouettes due to occlusions or shadows is introduced and it is shown that this approach has a great ability to estimate human poses from polluted silhouette with small computational burden.
Abstract
In this paper, we introduce a framework of human pose estimation from polluted silhouettes due to occlusions or shadows. Since the body pose (and configuration) can be estimated by partial components of the silhouette, a robust statistical method is applied to extract useful information from these components. In this method a Gaussian Process model is used to create each sub-manifold corresponding to the component of input data in advance. A sub-manifold voting strategy is then applied to infer the pose structure based on these sub-manifolds. Experiments show that our approach has a great ability to estimate human poses from polluted silhouettes with small computational burden.
Keywords
Computer ScienceDecision SciencesEngineering
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.
Gaussian Process Latent Variable Models for Visualisation of High Dimensional Data
684 Citations2003Neil D. Lawrence
A new underlying probabilistic model for principal component analysis (PCA) is introduced that shows that if the prior's covariance function constrains the mappings to be linear the model is equivalent to PCA, and is extended by considering less restrictive covariance functions which allow non-linear mappings.
ACM Transactions on GraphicsStyle-based inverse kinematics
647 Citations2004Keith Grochow, Steven Martin +2 more
Inferring 3D body pose from silhouettes using activity manifold learning
415 Citations2004Ahmed Elgammal, Chan-Su Lee
This work aims to infer 3D body pose directly from human silhouettes by explicitly learning view-based representations of activity manifolds as well as learning mapping functions between such central representations and both the visual input space and the 3DBody pose space.
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.
Estimating 3D hand pose from a cluttered image
385 Citations2003Vassilis Athitsos, Stan Sclaroff
A method is proposed that can generate a ranked list of plausible three-dimensional hand configurations that best match an input image that can function in the presence of clutter, thanks to two novel clutter-tolerant indexing methods.
Covariance scaled sampling for monocular 3D body tracking
293 Citations2005Cristian Sminchisescu, Bill Triggs
Experiments on some challenging monocular sequences show that robust cost modelling, joint and self-intersection constraints, and informed sampling are all essential for reliable monocular 3D body tracking.
Inferring 3D structure with a statistical image-based shape model
195 Citations2003Grauman, Shakhnarovich +1 more
The shape model enables accurate estimation of structure despite segmentation errors or missing views in the input silhouettes, and it works even with only a single input view, using a training set of thousands of pedestrian images generated from a synthetic model.
3D hand pose reconstruction using specialized mappings
192 Citations2002R. Rosales, Vassilis Athitsos +2 more
A system for recovering 3D hand pose from monocular color sequences is proposed, which employs a non-linear supervised learning framework, the specialized mappings architecture (SMA), to map image features to likely3D hand poses.
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.
Okapi-Chamfer matching for articulate object recognition
17 Citations2005Hanning Zhou, Thomas S. Huang
This paper proposes to improve the efficiency of articulated object recognition by an Okapi-Chamfer matching algorithm based on the inverted index technique, and test the system with both synthesized and real world images.
Cyclic articulated human motion tracking by sequential ancestral simulation
12 Citations2004Cheng Chang, Rashid Ansari +1 more
This paper presents a novel technique for tracking cyclic human motion based on decomposing complex cyclic motion into components and maintaining coupling between components and shows that the approach handles severe self-occlusion and foreign body occlusion with improved accuracy and efficiency.