文件列表:

Wavelet-based pattern analysis\templar\1-D\apology.txt (116, 2001-06-06)
Wavelet-based pattern analysis\templar\2-D\atomic_rep.m (657, 2001-07-20)
Wavelet-based pattern analysis\templar\2-D\classify.m (429, 2001-07-20)
Wavelet-based pattern analysis\templar\2-D\example1.m (2872, 2001-07-20)
Wavelet-based pattern analysis\templar\2-D\example2.m (3403, 2001-07-20)
Wavelet-based pattern analysis\templar\2-D\example_cl.m (1353, 2001-07-20)
Wavelet-based pattern analysis\templar\2-D\example_vid.m (2894, 2001-07-20)
Wavelet-based pattern analysis\templar\2-D\likelihood.m (1691, 2001-07-20)
Wavelet-based pattern analysis\templar\2-D\mostlikely.m (410, 2001-07-20)
Wavelet-based pattern analysis\templar\2-D\templar.m (7394, 2001-07-20)
Wavelet-based pattern analysis\templar\2-D\tools\bw.m (119, 2001-07-02)
Wavelet-based pattern analysis\templar\2-D\tools\circshift.m (944, 2003-11-07)
Wavelet-based pattern analysis\templar\2-D\tools\cp_rotate.m (511, 2001-06-01)
Wavelet-based pattern analysis\templar\2-D\tools\displayimagesc.m (118, 2000-04-18)
Wavelet-based pattern analysis\templar\2-D\tools\display_iterances.m (229, 2001-07-20)
Wavelet-based pattern analysis\templar\2-D\tools\display_registered_data.m (385, 2001-07-20)
Wavelet-based pattern analysis\templar\2-D\tools\display_stats.m (401, 2001-07-20)
Wavelet-based pattern analysis\templar\2-D\tools\display_synth_data.m (274, 2001-07-20)
Wavelet-based pattern analysis\templar\2-D\tools\display_template.m (490, 2001-07-20)
Wavelet-based pattern analysis\templar\2-D\tools\display_training_data.m (366, 2001-07-20)
Wavelet-based pattern analysis\templar\2-D\tools\eval_pll.m (493, 2001-07-20)
Wavelet-based pattern analysis\templar\2-D\tools\generate_scope.m (1035, 2001-05-22)
Wavelet-based pattern analysis\templar\2-D\tools\gen_rings.m (574, 2001-06-05)
Wavelet-based pattern analysis\templar\2-D\tools\gen_ri_dat.m (1106, 2001-06-05)
Wavelet-based pattern analysis\templar\2-D\tools\gen_squares.m (457, 2001-06-01)
Wavelet-based pattern analysis\templar\2-D\tools\gen_sq_dat.m (1108, 2001-06-05)
Wavelet-based pattern analysis\templar\2-D\tools\gen_tools.m (236, 2001-05-21)
Wavelet-based pattern analysis\templar\2-D\tools\gen_video.m (804, 2001-07-02)
Wavelet-based pattern analysis\templar\2-D\tools\hcircshift.m (438, 2001-06-01)
Wavelet-based pattern analysis\templar\2-D\tools\hsymshift.m (485, 2001-06-01)
Wavelet-based pattern analysis\templar\2-D\tools\init_wavelet_data.m (235, 2001-06-05)
Wavelet-based pattern analysis\templar\2-D\tools\inv_transform.m (88, 2001-05-22)
Wavelet-based pattern analysis\templar\2-D\tools\load_data.m (768, 2001-05-24)
Wavelet-based pattern analysis\templar\2-D\tools\ri_dat.mat (17968, 2001-06-05)
Wavelet-based pattern analysis\templar\2-D\tools\rot.m (224, 2001-06-01)
Wavelet-based pattern analysis\templar\2-D\tools\spatial_mean.m (95, 2001-06-07)
Wavelet-based pattern analysis\templar\2-D\tools\spatial_var.m (256, 2001-06-07)
Wavelet-based pattern analysis\templar\2-D\tools\sq_dat.mat (17968, 2001-06-05)
Wavelet-based pattern analysis\templar\2-D\tools\synthesize.m (1021, 2001-06-05)
Wavelet-based pattern analysis\templar\2-D\tools\transform.m (81, 2001-05-22)
... ...

Author of software: Clay Scott, cscott@rice.edu Based on: Clayton Scott and Robert Nowak, Template learning from atomic representations: A wavelet-based approach to pattern analysis, Workshop on SCTV, 2001, available at http://www.dsp.rice.edu/publications. %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Directory structure templar | |______1-D | | | |_______tools | |______2-D | |_______tools %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Description of directories 1-D - contains routines for implementing the learning and classification algorithms on 1-dimensional signals tools - various subroutines 2-D - contains routines for implementing the learning and classification algorithms on 2-dimensional signals tools - various subroutines %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Description of M-files in 1-D/2-D directories templar.m - given some training data, and a specified range of transformations, learn a template. type 'help templar' for more info. likelihood.m - given a template and a test signal/image, calculate the most likely transformation mapping the template to the signal/image, and return the transformation and the likelihood value. update_marginals.m - compute the most likely values for the means and variances of the significant coefficients, and the common variance of the insiginificant coefficients. update_states.m - update which wavelet coefficients are significant by maximizing the penalized log-likelihood update_transforms.m - for each training image, determine the most likely transformation mapping the template to that image classify.m - given some templates, classify a batch of test signals/images. see 2-D/example_cl.m mostlikely.m - determine which template best matches a given signal/image atomic_rep.m - atomic representation to be used (default: full haar wavelet transform). dont' forget to change inverse, too, if you make a change! %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Description of M-files in tools directories display_template.m - displays the mean, spatial variance, and a map of the significant states of the final template, and a synthesized image. display_training_data.m - displays training data. display_registered_data.m - displays training data after applying the inverse of the transform leaned by TEMPLAR synthesize.m - use template to synthesize T images. display_synth_data.m - Display the data synthesized by the previous command eval_pll.m - evaluate the penalized log-likelihood at any given configuration for the template and transforms load_data.m - a convenient program for loading training data. type "help load_data" for instructions. display_stats.m - display info about TEMPLAR while it's running. type "help templar for more %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Usage If you are using the two-dimenstional version, make sure your path includes \2-D \2-D\tools but does not include \1-D \1-D\tools. and vice-versa if using the 1-d version. Now look at files named example*.m for some simple examples of how to use the software. Be sure to modify atomic_rep.m before you begin, to specify which atomic rep you'll be using. (default: haar wavelet) Contact Clay Scott at cscott@rice.edu with questions.

近期下载者：

相关文件：

评论：[我要评论] [举报此文件]

收藏者：