L1-Homotopy-ALM

所属分类:matlab编程
开发工具:matlab
文件大小:77KB
下载次数:155
上传日期:2011-05-25 17:24:54
上 传 者wq319
说明:  基于稀疏表示的人脸识别,里面有9种求1范数的方法
(Face recognition based on sparse representation, there are nine kinds of seeking a method of norm)

文件列表:
L1_code (0, 2011-03-07)
L1_code\.DS_Store (6148, 2011-03-07)
__MACOSX (0, 2011-03-07)
__MACOSX\L1_code (0, 2011-03-07)
__MACOSX\L1_code\._.DS_Store (82, 2011-03-07)
L1_code\.svn (0, 2011-03-04)
L1_code\.svn\entries (882, 2011-03-04)
L1_code\.svn\prop-base (0, 2011-03-04)
L1_code\.svn\prop-base\nvmex.svn-base (30, 2011-03-04)
L1_code\.svn\prop-base\nvopts.sh.svn-base (30, 2011-03-04)
L1_code\.svn\props (0, 2011-03-04)
L1_code\.svn\text-base (0, 2011-03-04)
L1_code\.svn\text-base\Makefile.svn-base (1609, 2011-03-04)
L1_code\.svn\text-base\nvmex.svn-base (46065, 2011-03-04)
L1_code\.svn\text-base\nvopts.sh.svn-base (7922, 2011-03-04)
L1_code\.svn\text-base\SolveDALM_fast.cu.svn-base (11543, 2011-03-04)
L1_code\.svn\tmp (0, 2011-03-04)
L1_code\.svn\tmp\prop-base (0, 2011-03-04)
L1_code\.svn\tmp\props (0, 2011-03-04)
L1_code\.svn\tmp\text-base (0, 2011-03-04)
L1_code\dalm (0, 2011-03-07)
L1_code\dalm\fast (0, 2011-03-07)
L1_code\dalm\fast\Makefile (357, 2011-03-07)
__MACOSX\L1_code\dalm (0, 2011-03-07)
__MACOSX\L1_code\dalm\fast (0, 2011-03-07)
__MACOSX\L1_code\dalm\fast\._Makefile (186, 2011-03-07)
L1_code\dalm\fast\SolveDALM_fast.h (1301, 2011-03-07)
__MACOSX\L1_code\dalm\fast\._SolveDALM_fast.h (184, 2011-03-07)
L1_code\dalm\fast\SolveDALM_fast.m (4044, 2011-03-04)
__MACOSX\L1_code\dalm\fast\._SolveDALM_fast.m (187, 2011-03-04)
L1_code\dalm\fast\SolveDALM_fast_mex.cpp (7417, 2011-03-07)
__MACOSX\L1_code\dalm\fast\._SolveDALM_fast_mex.cpp (186, 2011-03-07)
L1_code\dalm\fast\test_fast.m (2249, 2011-03-06)
__MACOSX\L1_code\dalm\fast\._test_fast.m (186, 2011-03-06)
L1_code\dalm\fast\test_random.cpp (4111, 2011-03-07)
__MACOSX\L1_code\dalm\fast\._test_random.cpp (188, 2011-03-07)
L1_code\dalm\regular (0, 2011-03-07)
L1_code\dalm\regular\Makefile (342, 2011-03-07)
__MACOSX\L1_code\dalm\regular (0, 2011-03-07)
__MACOSX\L1_code\dalm\regular\._Makefile (186, 2011-03-07)
... ...

/*% README % Copyright 2010. The Regents of the University of California (Regents). % All Rights Reserved. Contact The Office of Technology Licensing, % UC Berkeley, 2150 Shattuck Avenue, Suite 510, Berkeley, CA 94720-1620, % (510) ***3-7201, for commercial licensing opportunities. % Created by Victor Shia, Allen Y. Yang, Department of EECS, University of California, % Berkeley. % IN NO EVENT SHALL REGENTS BE LIABLE TO ANY PARTY FOR DIRECT, INDIRECT, % SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES, INCLUDING LOST PROFITS, % ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN IF % REGENTS HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. % REGENTS SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING, BUT NOT LIMITED % TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A % PARTICULAR PURPOSE. THE SOFTWARE AND ACCOMPANYING DOCUMENTATION, IF ANY, % PROVIDED HEREUNDER IS PROVIDED "AS IS". REGENTS HAS NO OBLIGATION TO % PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS. */ README file for DALM/PALM and Homotopy l1-minimization code by Victor Shia, Allen Yang ================================ Homotopy and DALM Code solves the following problem: min ||x||_1 s.t. b = Ax There are two versions of the DALM code, fast and regular (matlab files included). PALM Code solves the following problems: min ||x||_1 + ||e||_1 s.t. b = Ax+e ================================ Two methods to make: (1) In matlab: mex -I -lmwblas -lmwlapack -DCOMPILE_MEX -DBLAS_IMPLEMENTATION_MATLAB -DCOMPILE_MEX (2) Using g++ (in the Makefile) Currently, there exist wrappers for the BLAS/LAPACK libraries for the ACML and MATLAB libraries. to use them, in the Makefile, add -DBLAS_IMPLEMENTATION_ to the compile commands Be sure to also include the header and library files for the corresponding libraries. Notes: (1) MATLAB libraries use a subset of Intel MKL's libraries, which are optimized for Intel Processors and can utilize multiple cores, unlike ACML libraries. I didn't write wrappers for the MKL libraries, but to increase performance, the user may want to use Intel MKL's libraries. (2) ACML libraries do NOT exist for Mac OSX ================================ test.m, test_random.cpp are just scripts to verify the output of the code in matlab and in C ================================ Email of Authors: Allen Yang: yang@eecs.berkeley.edu Victor Shia: vshia@eecs.berkeley.edu

近期下载者

相关文件


收藏者