~speedprog/mtg/mtg_card_detector.git - Our Gitblit

speedprog / mtg/mtg_card_detector

first | « prev | next »

2015-03-30

sides of box instead of coords

diff | tree

2015-03-30

diff | tree

2015-03-28

We do our OWN resizing!

diff | tree

2015-03-28

dropout probably ok

diff | tree

2015-03-27

Fixed im2col mistake >< face#palm

diff | tree

2015-03-25

smaller gridsize in bias

diff | tree

2015-03-24

stuff changed probably

diff | tree

2015-03-23

going to break stuff

diff | tree

2015-03-22

Subdivisions for batches

diff | tree

2015-03-21

using caffe's im2col, it's so much better\!

diff | tree

2015-03-21

gonna change im2col

diff | tree

2015-03-12

refactoring and added DARK ZONE

diff | tree

2015-03-08

renamed sigmoid to logistic

diff | tree

2015-03-08

detection layer fixed

diff | tree

2015-03-06

Split commands into different files

diff | tree

2015-03-06

going to split things up

diff | tree

2015-03-04

Big changes to detection

diff | tree

2015-02-24

diff | tree

2015-02-11

Detection is back, baby\!

diff | tree

2015-02-09

Generalizing conv layer so deconv is easier

diff | tree

2015-02-09

doing some testing w/o rand

diff | tree

2015-02-07

saving weight files as binaries, hell yeah

diff | tree

2015-02-04

Stable place to commit

diff | tree

2015-01-31

idk, probably something changed

diff | tree

2015-01-27

New name: Darknet

diff | tree

2015-01-27

Bias updates bug fix

diff | tree

2015-01-23

diff | tree

2015-01-20

I am so done with opencl, switching to cuda

diff | tree

2015-01-20

fast sort of working

diff | tree

2015-01-14

Stable, needs to be way faster

diff | tree

2015-01-13

diff | tree

2014-12-28

Need to fix line reads

diff | tree

2014-12-22

some fixes, some other experiments

diff | tree

2014-12-18

diff | tree

2014-12-18

diff | tree

2014-12-18

diff | tree

2014-12-18

changed nist testing

diff | tree

2014-12-18

diff | tree

2014-12-16

diff | tree

2014-12-16

lots of cleaning

diff | tree

2014-12-16

diff | tree

2014-12-13

fixed dropout ><

diff | tree

2014-12-11

Better imagenet distributed training

diff | tree

2014-12-08

diff | tree

2014-12-08

Fixed race condition in server

diff | tree

2014-12-08

diff | tree

2014-12-08

Some fixes to momentum

diff | tree

2014-12-07

Distributed training

diff | tree

2014-12-04

Detection good, split up col images

diff | tree

2014-12-03

Starting on server/client

diff | tree

first | « prev | next »