Dirichlet process mixture models on symmetric positive definite matrices for appearance clustering in video surveillance applications

Keyphrases

• Clustering Algorithm 100%
• Symmetric Positive Definite Matrix 100%
• Dirichlet Process Mixture Model 100%
• Video Surveillance Application 100%
• K-means 66%
• Computer Vision 33%
• Data Capture 33%
• Face Recognition 33%
• Bayesian Framework 33%
• Superior Performance 33%
• Mixture Model 33%
• Covariance Matrix 33%
• Unsupervised Clustering 33%
• Number of Components 33%
• Clustering Model 33%
• Divergence Measure 33%
• Riemannian Metric 33%
• Robustness to Noise 33%
• Frobenius Norm 33%
• Multivariate Data 33%
• Distance Measure 33%
• LogDet Divergence 33%
• Inverse Wishart 33%
• Feature Descriptor 33%
• Feature Correlation 33%
• Conjugate Pair 33%
• Euclidean Measure 33%
• Soft Clustering 33%
• Euclidean Geometry 33%
• Log-Euclidean 33%
• Norm-Euclidean 33%

Computer Science

• Video Surveillance 100%
• Dirichlet Process 100%
• Positive Definite Matrix 100%
• Clustering Algorithm 100%
• Superior Performance 33%
• Face Recognition 33%
• feature descriptor 33%
• Covariance Matrix 33%
• Bayesian Framework 33%
• Multivariate Data 33%
• Distance Measure 33%
• Euclidean Geometry 33%
• Computer Vision 33%
• Unsupervised Clustering 33%