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: