fische_r
|
ead3eba3aa
|
added full temporal mean as function, but not activated in the prepare()
|
2023-02-13 16:29:08 +01:00 |
|
Fischer Robert
|
84908c3795
|
added pixel coordinate as feature
|
2023-01-18 15:22:32 +01:00 |
|
Fischer Robert
|
ef919a2db3
|
Merge branch 'master' of https://gitlab.psi.ch/fische_r/pytrainseg
|
2023-01-17 14:04:52 +01:00 |
|
Fischer Robert
|
e42052328c
|
local copy
|
2023-01-17 14:04:33 +01:00 |
|
Fischer Robert
|
8c0cc68e57
|
TODO: store git sha
|
2022-10-13 15:13:18 +02:00 |
|
Fischer Robert
|
13f6fbfa2a
|
worked around dask crashing/zombieing when fetching results
|
2022-09-26 13:56:07 +02:00 |
|
Fischer Robert
|
193ca64faf
|
some bug fixes, but not yet there
|
2022-09-16 14:53:22 +02:00 |
|
Fischer Robert
|
942bad4eba
|
training now working with very large datasets
|
2022-09-06 16:03:02 +02:00 |
|
Fischer Robert
|
0633b8a129
|
wpatched: when lazy cannot use originalà as reference for gauss with sig 0
|
2022-09-05 11:02:58 +02:00 |
|
Fischer Robert
|
45f8074bbd
|
making it fully lazy
|
2022-09-01 17:33:22 +02:00 |
|
Fischer Robert
|
e77035eef5
|
segmenting lines for wood
|
2022-08-24 15:50:23 +02:00 |
|
Fischer Robert
|
ca7f6b0677
|
worked again with wood
|
2022-08-24 15:08:23 +02:00 |
|
Fischer Robert
|
966af5bfaf
|
update
|
2022-08-22 08:55:49 +02:00 |
|
Fischer Robert
|
d44ab4c143
|
looks ok
|
2022-08-18 17:12:37 +02:00 |
|
Fischer Robert
|
c350295d5c
|
works neatly
|
2022-08-18 16:20:56 +02:00 |
|
Fischer Robert
|
44abe23af1
|
I think I have all
|
2022-08-18 14:57:21 +02:00 |
|
Fischer Robert
|
6630acfedc
|
got it running, now allow clf savinf and full segmentation
|
2022-08-18 11:40:45 +02:00 |
|
fische_r
|
c48bdb1174
|
put all together, but ipycanvas not working on laptop
|
2022-08-18 11:06:46 +02:00 |
|
fische_r
|
c8e40e4510
|
started training part
|
2022-08-17 17:10:25 +02:00 |
|
Fischer Robert
|
8b9b292458
|
add notebook to test dask
|
2022-08-17 14:08:54 +02:00 |
|
fische_r
|
e9a9442556
|
bug with range elemnt in Hessian
|
2022-08-17 14:06:10 +02:00 |
|
fische_r
|
cdd74c6d62
|
messed up with gauss 4d dict name
|
2022-08-17 14:04:51 +02:00 |
|
fische_r
|
88fa3b8848
|
forgot combinations in gauss diff
|
2022-08-17 14:03:25 +02:00 |
|
fische_r
|
5ae07895f3
|
check diff gauss later
|
2022-08-17 14:03:07 +02:00 |
|
fische_r
|
101ddff092
|
forgot combinations in gauss diff
|
2022-08-17 14:01:30 +02:00 |
|
fische_r
|
fe3648e6a6
|
inlcuded all? filters?
|
2022-08-17 13:58:28 +02:00 |
|
fische_r
|
e8012344f7
|
inlcuded all? filters?
|
2022-08-17 13:56:11 +02:00 |
|
fische_r
|
2cbd888c89
|
saving ok, filling up filters
|
2022-08-17 10:46:18 +02:00 |
|
fische_r
|
86dc75b727
|
fixed some stuff
|
2022-08-17 08:59:14 +02:00 |
|
fische_r
|
a38047b357
|
removed dask rename, find other way
|
2022-08-16 16:46:13 +02:00 |
|
fische_r
|
7c3a99218c
|
added local dask cluster
|
2022-08-16 16:14:36 +02:00 |
|
fische_r
|
f8eba110cc
|
ready to test?!
|
2022-08-16 15:59:16 +02:00 |
|
fische_r
|
83f1698fcf
|
build into a class hopefully using an efficient dask tree
|
2022-08-16 15:40:44 +02:00 |
|
Fischer Robert
|
c48e48c6c5
|
added plotting
|
2022-08-15 16:17:07 +02:00 |
|
Fischer Robert
|
85e4f66d8f
|
test on full volume of jeremy's data, no data streaming yet necessary
|
2022-08-15 14:58:03 +02:00 |
|
Fischer Robert
|
dbf02fc03e
|
pretty decent for Jeremy's data
|
2022-08-12 10:41:53 +02:00 |
|
Fischer Robert
|
41ed977005
|
not yet working for wood (swelling), needs some sort of time space/correlation
|
2022-08-12 10:40:51 +02:00 |
|
Fischer Robert
|
94a0903624
|
TODO: function to load label images, clever filters
|
2022-08-11 17:26:46 +02:00 |
|
Fischer Robert
|
1710a8b5f5
|
works better now, inlcude average/max Hessian
|
2022-08-10 16:51:26 +02:00 |
|
Fischer Robert
|
636b3d195e
|
added anisotropic time filtering
|
2022-08-10 15:06:05 +02:00 |
|
Fischer Robert
|
5f9e341055
|
wood segmentation not good for fiber/water long time after filling
|
2022-08-10 10:47:32 +02:00 |
|
Fischer Robert
|
4e4a320246
|
very good on interlaces
|
2022-08-10 10:45:51 +02:00 |
|
Fischer Robert
|
06f71cdf1f
|
testing segmentation of yarns, try also on wood
|
2022-08-09 11:04:34 +02:00 |
|
Fischer Robert
|
035ffc472f
|
saving segmented data
|
2022-08-08 16:31:59 +02:00 |
|
Fischer Robert
|
cd4282ceac
|
Merge branch 'master' of https://gitlab.psi.ch/fische_r/pytrainseg
|
2022-08-08 16:27:43 +02:00 |
|
Fischer Robert
|
b16f4adbee
|
testing clf saving and application
|
2022-08-08 16:27:20 +02:00 |
|
Fischer Robert
|
7cf38a21ac
|
last saved version before quota issue
|
2022-08-08 15:28:34 +02:00 |
|
Fischer Robert
|
45bea73937
|
hopefully saved all changes from new WS maunally
|
2022-08-08 15:22:30 +02:00 |
|
Fischer Robert
|
ebac638269
|
inlcuded image gradient to features
|
2022-08-05 16:45:09 +02:00 |
|
Fischer Robert
|
da84c5924a
|
full workflow to be tested
|
2022-08-05 13:43:51 +02:00 |
|