Commits

Branch selector

 master

User selector

All users

Datepicker

All time

Commit History

Commits on Oct 15, 2015

• Merge pull request vlfeat#278 from suhangpro/patch-1

  Show description for dde29bc
  vedaldi

  committed
  dde29bc

Commits on Oct 13, 2015

• cnn_train_dag.m: better naming of losses

  vedaldi

  committed
  20717fe

• beta15 -> beta16

  vedaldi

  committed
  3980342

• vgg-face: adds to the pertained models

  vedaldi

  committed
  c61dc08

• cnn_imagenet_googlenet.m: fixes for latest version of downloadable model

  vedaldi

  committed
  280c030

Commits on Oct 12, 2015

• adds vl_nnnormalizelp.m

  vedaldi

  committed
  9cfcfed

• vl_testnn.m: works in a directory other than MATCONVNETROOT

  vedaldi

  committed
  108b198

• DagNN: print: uses display in Linux to visualize the generated PDF

  vedaldi

  committed
  b02f6a0

• doc: sync with latest changes to the DAG converter

  vedaldi

  committed
  4d40a19

• utils: DAG import script bugfixes

  vedaldi

  committed
  9308c75

Commits on Oct 11, 2015

• utils: improt dag: bugfixes when importing inner-product layer

  vedaldi

  committed
  a73c76f

• utils/import-caffe-dag.py: bugfix

  vedaldi

  committed
  7beee59

• dagnn.Crop: fixes vlfeat#288

  vedaldi

  committed
  66e93dd

• DagNN: add rudimentary support for displaying a graph using Graphviz and dot

  vedaldi

  committed
  830040a

Commits on Oct 8, 2015

• doc: site: make the link to FCN training code more prominent

  vedaldi

  committed
  6afeaea

• doc: site: adds a pointer to the FCN training code

  vedaldi

  committed
  de55e9d

Commits on Oct 6, 2015

• fix a bug causing error when the first layer has no params

  suhangpro

  committed
  95397b8

Commits on Oct 5, 2015

• vl_nndropout: clarifies documentation (thanks to Jasper; closes vlfeat#208)

  vedaldi

  committed
  c07d4c0

• doc: fixes typos

  vedaldi

  committed
  b52bd00

Commits on Oct 4, 2015

• doc: pretrained: fixes description of FCN-TVG

  vedaldi

  committed
  18a2886

• doc: corrects attribution of GoogLeNet pertained model

  vedaldi

  committed
  151cf66

• doc: cosmetic fix.

  vedaldi

  committed
  2a8540d

• doc: pretrained: adds TVG FCN model

  vedaldi

  committed
  17cb8d4

• doc: clarifies section on pooling geometry

  vedaldi

  committed
  6524cf2

• utils: remove obsolete import script for DAG

  vedaldi

  committed
  0cecf0e

• utils: remove obsolete import script for DAG

  vedaldi

  committed
  43ef633

• utils: improve import scripts for FCN and GoogLeNet

  vedaldi

  committed
  be4063d

Commits on Oct 3, 2015

• Merge branch 'master' of ssh://bitbucket.org/ovl/matconvnet

  vedaldi

  committed
  c45dc0a

• utils: import googlnet using the new script

  vedaldi

  committed
  4d94e84

• utils: import script: bugfixs

  vedaldi

  committed
  09cab7c

Commits on Oct 1, 2015

• Xcode: adds missing files

  vedaldi

  committed
  12673ee

• crop: minor fixes to the doc

  vedaldi

  committed
  7fb017b

• doc: small improvements

  vedaldi

  committed
  42ac2b1

• bnorm: cosmetic fixes

  vedaldi

  committed
  fd676c9

• doc: adds an explanation of how padding differs in MatConvNet pooling and Caffe pooling

  vedaldi

  committed
  0c1430a

Pagination

Previous

Next