Name
Abstract | ||
|---|---|---|
In this paper, a Bayesian mixture expert (BME) framework for the estimation of 3D human poses from two uncalibrated wide-baseline cameras is presented. The two cameras will reduce the ambiguities of the pose estimation greatly and is easy to implement. BME is learnt to conduct multimodal pose estimation regression. K-means algorithm considering Euclidean distance and maximum-value distance for the joint angle vector is used for the initial clustering in BME learning. This will give the better cluster results to separate the ambiguous poses into different experts. Also a weighted PCA is implemented in an expectation-maximization (EM) framework to learn the parameters of the BME. This can reduce the dimension of the training data more effectively compared with global PCA. The system is trained with synthesized silhouettes from motion capture data. The experimental results on synthesized and real images illustrate that our approach does not need precise camera calibration and can estimate the poses effectively |
Year | DOI | Venue |
|---|---|---|
2007 | 10.1109/WACV.2007.31 | WACV |
Keywords | Field | DocType |
bayesian mixture expert,expectation-maximisation algorithm,calibration,synthesized silhouettes,bme learning,maximum value distance,estimation regression,bayes methods,different expert,learning (artificial intelligence),euclidean distance,weighted pca,uncalibrated wide baseline cameras,motion capture data,pose estimation,bme learning clustering,expectation-maximization framework,synthesized silhouette,bayesian mixture expert framework,cameras,stereo cameras,maximum-value distance,human pose inference,multimodal pose estimation regression,global pca,stereo image processing,k-means algorithm,principal component analysis,3d human pose estimation,camera calibration,training data,joint angle vector,image motion analysis,expectation maximization,k means algorithm,learning artificial intelligence | Computer vision,Stereo cameras,Motion capture,Pattern recognition,Computer science,Inference,Euclidean distance,Pose,Camera resectioning,Artificial intelligence,Real image,Cluster analysis | Conference |
ISSN | ISBN | Citations |
1550-5790 E-ISBN : 0-7695-2794-9 | 0-7695-2794-9 | 1 |
PageRank | References | Authors |
0.35 | 9 | 2 |