A multiphase region-based framework for image segmentation based on least square method

Chen, G.; Meng, Xiangxu; Hu, Taotao; Guo, Xifeng; Liu, Lixiong; Zhang, Haiying

doi:10.1109/icip.2009.5413829

Cited by 1 publication

(1 citation statement)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recently, Knowles et al proposed a Dirichlet process variable clustering (DPVC) method by leveraging the correlation between variables and formulating the corresponding probabilistic model for non-parametric clustering (Knowles, Palla, and Ghahramani 2012). In general, typical clustering algorithms, such as K-means and GMM (Chen et al 2009), consider how similar entities (in terms of Euclidean distance) rather how correlated they are. Thus, DPVC can discover block diagonal covariance structures in data, and partition observed variables into sets of highly correlated variables for clustering.…”

Section: Related Workmentioning

confidence: 99%

Maximum Margin Dirichlet Process Mixtures for Clustering

Chen

Zhang

Xiong³

2016

AAAI

View full text Add to dashboard Cite

The Dirichlet process mixtures (DPM) can automatically infer the model complexity from data. Hence it has attracted significant attention recently, and is widely used for model selection and clustering. As a generative model, it generally requires prior base distribution to learn component parameters by maximizing posterior probability. In contrast, discriminative classifiers model the conditional probability directly, and have yielded better results than generative classifiers.In this paper, we propose a maximum margin Dirichlet process mixture for clustering, which is different from the traditional DPM for parameter modeling. Our model takes a discriminative clustering approach, by maximizing a conditional likelihood to estimate parameters. In particular, we take a EM-like algorithm by leveraging Gibbs sampling algorithm for inference, which in turn can be perfectly embedded in the online maximum margin learning procedure to update model parameters. We test our model and show comparative results over the traditional DPM and other nonparametric clustering approaches.

show abstract

Section: Related Workmentioning

confidence: 99%