Unsupervised image classification, segmentation, and enhancement using ICA mixture models

IEEE Transactions on Image Processing

Volumn 11, Issue 3, 2002, Pages 270-279

  Lee, Te Won ^a   Lewicki, Michael S ^b  

^a UNIVERSITY OF CALIFORNIA SAN DIEGO   (United States)

^b Carnegie Mellon University ^*  (United States)

Author keywords

Blind source separation; Denoising; Fill in missing data; Gaussian mixture model; Image coding; Independent component analysis; Maximum likelihood; Segmentation; Unsupervised classification

Indexed keywords

BLIND SOURCE SEPARATION; COMPUTER SIMULATION; IMAGE CODING; IMAGE ENHANCEMENT; IMAGE SEGMENTATION; INDEPENDENT COMPONENT ANALYSIS; MAXIMUM LIKELIHOOD ESTIMATION; PATTERN RECOGNITION;

DENOISING; GAUSSIAN MIXTURE MODEL; INDEPENDENT COMPONENT ANALYSIS MIXTURE MODEL; UNSUPERVISED IMAGE CLASSIFICATION;

LEARNING ALGORITHMS;

EID: 0036504424     PISSN: 10577149     EISSN: None     Source Type: Journal    
DOI: 10.1109/83.988960     Document Type: Article

References (39)

1. Blind separation of noisy mixtures: An EM algorithm for independent factor analysis
2. Blind signal separation in teleconferencing using the ica mixture model
3. An information-maximization approach to blind separation and blind deconvolution
4. The independent components' of natural scenes are edge filters
5. Mixture density networks
6. High quality document image compression with DJVU
7. Image compression via joint statistical characterization in the wavelet domain
8. Equivariant adaptive source separation
9. High-order contrast for independent component analysis
10. Mathematical models for local deterministic inpaintings
11. Independent component analysis-a new concept?
12. A nonparametric multi-scale statistical model for natural images
13. Texture synthesis by nonparametric sampling
14. What is the goal of sensory coding?
15. The EM algorithm for mixtures of factor analyzers
16. An alternative perspective on adaptive independent component analysis algorithms
17. A fast fixed-point algorithm for independent component analysis
18. Sparse code shrinkage: Denoising by nonlinear maximum likelihood estimation
19. Blind separation of sources, part I: An adaptive algorithm based on neuromimetic architecture
20. The generalized Gaussian mixture model
21. A unifying framework for independent component analysis
22. Independent component analysis using an extended infomax algorithm for mixed sub-Gaussian and super-Gaussian sources
23. The generalized Gaussian mixture model using ICA
24. Unsupervised classification with non-Gaussian mixture models using ICA
25. Unsupervised classification with non-Gaussian sources and automatic context switching in blind signal separation
26. Inferring sparse, overcomplete image codes using an efficient coding framework
27. Learning nonlinear overcomplete representations for efficient coding
28. A probabilistic framework for the adaptation and comparison of image codes
29. Learning overcomplete representations
30. A flexible prior for independent component analysis
31. Multi-band approaches for texture classification and segmentation
32. Emergence of simple-cell receptive field properties by learning a sparse code for natural images
33. Autoclass-a Bayesian approach to classification
34. Minimax entropy principle and its application to texture modeling