文件列表:

lk20_p2_release\lk20_common\hessian_weight.m (466, 2004-01-15)
lk20_p2_release\lk20_common\sd_update_weight.m (490, 2004-01-15)
lk20_p2_release\lk20_common\hessian.m (461, 2004-01-09)
lk20_p2_release\lk20_common\init_a.m (710, 2004-01-09)
lk20_p2_release\lk20_common\jacobian_a.m (535, 2004-01-09)
lk20_p2_release\lk20_common\quadtobox.m (1312, 2004-01-09)
lk20_p2_release\lk20_common\sd_images.m (541, 2004-01-09)
lk20_p2_release\lk20_common\sd_update.m (491, 2004-01-09)
lk20_p2_release\lk20_common\hessian_sub.m (683, 2004-01-15)
lk20_p2_release\lk20_common\verb_plot_a.m (648, 2004-01-09)
lk20_p2_release\lk20_common\robust_error.m (546, 2004-01-15)
lk20_p2_release\lk20_common\warp_a.m (771, 2004-01-09)
lk20_p2_release\lk20_common\compute_weight.m (465, 2004-01-15)
lk20_p2_release\lk20_common\occlude_image.m (1913, 2004-01-15)
lk20_p2_release\lk20_common\normalize_weight.m (236, 2004-01-15)
lk20_p2_release\lk20_common\sd_images_weight.m (533, 2004-01-15)
lk20_p2_release\lk20_common\compute_block_coords.m (928, 2004-01-15)
lk20_p2_release\lk20_common\compute_block_weight.m (856, 2004-01-15)
lk20_p2_release\lk20_common (0, 2007-08-24)
lk20_p2_release\lk20_p2\affine_ic_irls_h.m (2924, 2004-01-15)
lk20_p2_release\lk20_p2\affine_ic_irls.m (2923, 2004-01-15)
lk20_p2_release\lk20_p2\affine_ic.m (2727, 2004-01-15)
lk20_p2_release\lk20_p2\figures\cnvg_plot_robust.m (1735, 2004-01-16)
lk20_p2_release\lk20_p2\figures\lk20_p2style.m (483, 2004-01-16)
lk20_p2_release\lk20_p2\figures\div_plot_robust.m (1957, 2004-01-16)
lk20_p2_release\lk20_p2\figures (0, 2007-08-24)
lk20_p2_release\lk20_p2\run_weighted.m (2209, 2004-01-20)
lk20_p2_release\lk20_p2\data\takeo.mat (34128, 2004-01-09)
lk20_p2_release\lk20_p2\data\affine_pt_offset.mat (240200, 2004-01-09)
lk20_p2_release\lk20_p2\data\occl_black.mat (262336, 2004-01-09)
lk20_p2_release\lk20_p2\data\occl_sequence.mat (480200, 2004-01-09)
lk20_p2_release\lk20_p2\data\occl_avgGL.mat (262336, 2004-01-13)
lk20_p2_release\lk20_p2\data\occl_scenary.mat (497856, 2004-01-16)
lk20_p2_release\lk20_p2\data\occl_face.mat (181776, 2004-01-16)
lk20_p2_release\lk20_p2\data (0, 2007-08-24)
lk20_p2_release\lk20_p2\results (0, 2007-08-24)
lk20_p2_release\lk20_p2\run_robust.m (2933, 2004-01-20)
lk20_p2_release\lk20_p2\affine_ic_irls_spatial.m (3449, 2004-01-15)
lk20_p2_release\lk20_p2\test_robust.m (7586, 2004-01-15)
... ...

Lucas-Kanade 20 Years On: A Unifying Framework: Part 2 Simon Baker, Ralph Gross, Takahiro Ishikawa, and Iain Matthews The Robotics Institute Carnegie Mellon University Pittsburgh, PA 15213 USA Contact: simonb@cs.cmu.edu, rgross@cs.cmu.edu, taka@cs.cmu.edu, iainm@cs.cmu.edu --------------------------------------------------------------------------- URL: http://www.ri.cmu.edu/projects/project_515.html Bibtex reference: @TechReport{Baker_2003_4302, author = "Simon Baker, Ralph Gross, Takahiro Ishikawa, and Iain Matthews", title = "Lucas-Kanade 20 Years On: A Unifying Framework: Part 2", institution = "Robotics Institute, Carnegie Mellon University", year = "2003", key = "CMU-RI-TR-03-01", } --------------------------------------------------------------------------- MATLAB SOURCE CODE --------------------------------------------------------------------------- The source code is split into two directories, both need to be added to your Matlab search path: lk20_p2 The algorithms described in the paper and test scripts lk20_common Common code for all algorithms. lk20_p2/ -------- The file affine_ic.m contains the original inverse compositional algorithm (described in Section 2.2). The algorithm in affine_ic_weighted.m implements a weighted L2 norm (described in Section 3). Algorithms using a robust error function (Section 4) affine_ic_irls.m Inverse Compositional Iteratively Reweighted Least Squares (4.3) affine_ic_irls_h.m H approximation to the IRLS algorithm (4.4.1) affine_ic_irls_spatial.m Spatial coherence approximation to the IRLS algorithm (4.4.2) To run the perturbation experiments described in the paper examine: run_weighted.m Run comparison between affine_ic and affine_ic_weighted (3.2.1) run_robust.m Run comparison between affine_ic, affine_ic_irls, affine_ic_irls_h and affine_ic_irls_spatial (4.6.3) These files should be edited for the experimental parameters you would like to test. By default they run everything for a few tests. To replicate the experiments in the paper run each algorithm for 100 convergence tests and 1000 frequency of convergence tests. This will take some time! lk20_p2/data/ ------------- The data directory contains precomputed random perturbation data (for experimental consistency) and the images and template definitions. lk20_p2/figures/ ---------------- This directory contains matlab scripts for plotting your results and creating graphs similar to those shown in the paper.

近期下载者：

相关文件：

评论：[我要评论] [举报此文件]

收藏者：