1 / 23

Applications of Independent Component Analysis

Applications of Independent Component Analysis. Terrence Sejnowski. Computational Neurobiology Laboratory The Salk Institute. PCA finds the directions of maximum variance ICA finds the directions of maximum independence. Principle: Maximize Information .

israel
Download Presentation

Applications of Independent Component Analysis

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Applications of Independent Component Analysis Terrence Sejnowski Computational Neurobiology Laboratory The Salk Institute

  2. PCA finds the directions of maximum variance ICA finds the directions of maximum independence

  3. Principle: Maximize Information • ICA produces brain-like visual filters for natural images. • A: ICA does this -- it maximizes joint entropy & minimizes mutual information between output channels (Bell & Sejnowski, 1995). • Q: How to extract maximum information from multiple visual channels? Set of 144 ICA filters

  4. Example: Audio decomposition Perform ICA Mic 1 Mic 2 Mic 3 Mic 4 Terry Scott Te-Won Tzyy-Ping Play Mixtures Play Components

  5. Sound source separation Image processing Sonar target identification Underwater communications Wireless communications Brain wave analysis (EEG) Brain imaging (fMRI) ICA Applications

  6. Recordings in real environments Separation of Music & Speech Experiment-Setup: - office room (5m x 4m) - two distant talking mics - 16kHz sampling rate 60cm 40cm

  7. Learning Image Features

  8. Learning Image Features

  9. Automatic Image Segmentation

  10. Barcode Classification Matrix Linear Postal

  11. Learned ICA Output Filters Matrix Linear Postal

  12. Barcode Classification Results Classifying 4 data sets: linear, postal, matrix, junk

  13. Image De-noising

  14. Filling in missing data

  15. ICA applied to Brainwaves An EEG recording consists of activity arising from many brain and extra-brain processes

  16. Eye movement Muscle activity

  17. WHAT ARE THE INDEPENDENT COMPONENTS OF BRAIN IMAGING? Task-related activations Arousal Measured Signal Physiologic Pulsations Machine Noise ?

  18. Functional Brain Imaging • Functional magnetic resonance imaging (fMRI) data are noisy and complex. • ICA identifies concurrent hemodynamic processes. • Does not require a priori knowledge of time courses or spatial distributions.

  19. Contact: terry@salk.edu ICA-2001: http://www.ica2001.org

More Related