A guide to independent component analysis (ICA)

• To start learning ICA, I would suggest the next paper. It's a simple monograph of less than three pages, describing the principle of ICA very nicely.
  • J.-F. Cardoso. Infomax and maximum likelihood for source separation. IEEE Letters on Signal Processing, 4: 112-114, 1997. [preprint]
    
  
  
• There are many ICA codes available on the web, and what I found most useful is Tony Bell's code. The code is simple and the algorithm is powerful, especially for a large-scale application. [code] [Tony Bell's ICA page]
  
  
• Three most important papers about the relation between ICA and visual coding are as follows.
  
  • B. A. Olshausen & D. J. Field. Sparse coding with an overcomplete basis set: A strategy employed by V1? Vision Research, 37: 3311-3325, 1997. [reprint]
    
  • A. J. Bell & T. J. Sejnowski. The "independent components" of natural scenes are edge filters. Vision Research, 37: 3327-3338, 1997. [reprint]
    
  • M. S. Lewicki & B. A. Olshausen. Probabilistic framework for the adaptation and comparison of image codes. JOSA, 16: 1587-1601, 1999. [reprint]
    
  
  
• You can find an active and up-to-date research on ICA at Hyvarinen's lab. There are many on-line papers, codes, links, and also an introduction to ICA. [Aapo Hyvarinen's page]