Skip to content
Commits on Jan 15, 2015
  1. @vedaldi

    Windows: defaults to use MATLAB 2010B for compilation

    vedaldi committed
    This is done in order to create redistributable binaries that are more compatible.
    However, there is no need to use an old MATLAB to compile the library.
  2. @vedaldi

    [bugfix] make/matlab.mak: fixes a bug causing the compilation to fail…

    vedaldi committed
    … under Linux and MATLAB R2010
  3. @vedaldi

    README.md: -> 0.9.20

    vedaldi committed
  4. @vedaldi
  5. @vedaldi
  6. @vedaldi
  7. @vedaldi
Commits on Jan 14, 2015
  1. @vedaldi
  2. @vedaldi
  3. @vedaldi
  4. @vedaldi
  5. @vedaldi

    Visual C: upgrade to VC 2010

    vedaldi committed
  6. @vedaldi

    Makefile.mak: -> 0.9.20

    vedaldi committed
  7. @vedaldi

    Xcode: sync

    vedaldi committed
  8. @vedaldi

    doc: adds 0.9.20 notes

    vedaldi committed
  9. @davnov134 @vedaldi
  10. @vedaldi

    generic.c: 0.9.19 -> 0.9.20

    vedaldi committed
  11. @vedaldi

    [enhancement] gmm: improves the algorithm used to restart empty Gauss…

    vedaldi committed
    …ian components
    
    This patch correctes the criterion used to decide which Gaussian components should be divided in order to restart a dead Gaussian component. The previous version was using an incorrect derivation of the math.
  12. @vedaldi
  13. @vedaldi
  14. @vedaldi

    hog.c: improves the comments in the code explaining in more detail th…

    vedaldi committed
    …e block structure in HOG
    
    hog.c: fix whitespace errors
Commits on Oct 31, 2014
  1. @vedaldi
Commits on Oct 13, 2014
  1. @vedaldi
  2. @vedaldi

    hog.c: fix a bug in the bilinear interpolated variant of the descriptor

    vedaldi committed
    Fortunately this variant is seldom used in practice.
Commits on Oct 4, 2014
  1. @vedaldi
  2. @vedaldi
Commits on Oct 2, 2014
  1. @pmoulon @vedaldi

    vl_covdet.m, covdet.c: fix typos

    pmoulon committed with vedaldi
  2. @vedaldi
  3. @pmoulon @vedaldi

    Update vl_covdet.c

    pmoulon committed with vedaldi
    Fix a memory leak.
  4. @vedaldi

    Merge pull request #55 from achalddave/master

    vedaldi committed
    vl_kmeans.m: update vl_kmeans doc to include MaxNumIterations
  5. @vedaldi

    Merge pull request #58 from achalddave/remove-minengvar-prints

    vedaldi committed
    vl_kmeans.c: remove unnecessary kmeans MinEnergyVar prints
  6. @vedaldi

    Merge pull request #56 from pmoulon/patch-3

    vedaldi committed
    generic.c: fix typos
  7. @vedaldi

    Merge pull request #57 from pmoulon/patch-4

    vedaldi committed
    vl_liop.m: fix typo
  8. @vedaldi

    vlad.c: fixes doc

    vedaldi committed
Commits on Sep 11, 2014
  1. @vedaldi
Something went wrong with that request. Please try again.