Luca Carlone
|
b51a8380c4
|
added boost include
|
2013-10-12 16:21:15 +00:00 |
Frank Dellaert
|
365d2d2d03
|
A retract test
|
2013-10-12 16:17:27 +00:00 |
Duy-Nguyen Ta
|
8be7363a01
|
refactor and add more comments
|
2013-10-12 15:27:37 +00:00 |
Frank Dellaert
|
1f293294fd
|
Some refactoring in Cal3Bundler
|
2013-10-12 15:04:26 +00:00 |
Frank Dellaert
|
a423849afe
|
Comments and formatting, plus some added tests on Cal3Bundler
|
2013-10-12 14:31:18 +00:00 |
Frank Dellaert
|
7d0674fe4b
|
Inlined derivatives in Cal3DS2, gets rid of a lot of duplicated calculations.
|
2013-10-12 07:07:08 +00:00 |
Frank Dellaert
|
31beb98a36
|
The copy/paste from CalibratedCamera (can we not avoid this?) had the side effect of not benefiting from the collapsing derivatives there. I added this in Pinhole now, which should shave off quite a bit of time in the derivative calculation.
|
2013-10-12 06:31:28 +00:00 |
Frank Dellaert
|
1a20272fa2
|
Reformatted with new style file, renamed some derivatives to Dcal, Dpose, Dpoint etc.
|
2013-10-12 06:02:16 +00:00 |
Duy-Nguyen Ta
|
b161a106c7
|
loopy belief in progress
|
2013-10-12 05:49:11 +00:00 |
Frank Dellaert
|
830ff38b80
|
Timing script for PinholeCamera
|
2013-10-12 05:18:40 +00:00 |
Frank Dellaert
|
fa140cb0a3
|
Reformatted with new style file, renamed some derivatives to Dcal, Dpose, Dpoint etc. Small edit to not compute Dcal if not asked.
|
2013-10-12 05:17:01 +00:00 |
Frank Dellaert
|
71466e8476
|
Added effing *unit tests* for PinholeCamera. It is pretty terrible that there were none !
|
2013-10-12 05:15:01 +00:00 |
Frank Dellaert
|
752a9877c5
|
Reformatted with new style file, renamed some derivatives to Dcal, Dpose, Dpoint etc.
|
2013-10-12 05:13:36 +00:00 |
Richard Roberts
|
ca9caf6a66
|
Hopefully fixed exception handling in TBB (again)
|
2013-10-11 22:17:15 +00:00 |
Duy-Nguyen Ta
|
35924f6055
|
disable printing
|
2013-10-11 21:59:36 +00:00 |
Duy-Nguyen Ta
|
2570492fa6
|
disable printing
|
2013-10-11 21:58:09 +00:00 |
Duy-Nguyen Ta
|
a5f1d68099
|
fix scheduling examples by adding the default key index ordering in Scheduler.
|
2013-10-11 21:55:31 +00:00 |
Duy-Nguyen Ta
|
00f109e88f
|
fix warnings
|
2013-10-11 20:26:50 +00:00 |
Duy-Nguyen Ta
|
19d64460ba
|
check in fail DiscreteBayesNet::sample unittest.
|
2013-10-11 19:07:16 +00:00 |
Duy-Nguyen Ta
|
b43ff9d377
|
disable printing in test
|
2013-10-11 19:05:43 +00:00 |
Duy-Nguyen Ta
|
8add3de3c2
|
fix compilation errors in unstable discrete examples
|
2013-10-11 19:01:12 +00:00 |
Duy-Nguyen Ta
|
1531abd285
|
fix testCSP by using the default key ordering.
|
2013-10-11 17:48:44 +00:00 |
Duy-Nguyen Ta
|
bf11f93cee
|
revive unstable discrete. testCSP still fails, but looks like the solution is valid. See comments in file.
|
2013-10-11 17:42:30 +00:00 |
Zsolt Kira
|
9639660685
|
Added GenericProjectionFactorCreator. Both smart and generic projection factors now work (again) in batch mode. Incremental not tested yet.
|
2013-10-11 01:59:36 +00:00 |
Duy-Nguyen Ta
|
6789c5080c
|
comments request!!!
|
2013-10-10 22:02:57 +00:00 |
Duy-Nguyen Ta
|
a2138084db
|
move print to cpp to wrap properly in matlab
|
2013-10-10 22:01:48 +00:00 |
Duy-Nguyen Ta
|
9aede467d2
|
revive discrete, except testDiscreteBayesNet::sample, and the whole testDiscreteBayesTree, currently disabled. It's still a mess!!!
|
2013-10-10 21:59:49 +00:00 |
Zsolt Kira
|
1a8a670870
|
Fixed bugs in new version since creator class is used: Get number of landmarks from creator object,
replace ensureHasOrdering with direct COLAMD call.
Current Smart Projection factor version works in batch mode
|
2013-10-10 20:21:40 +00:00 |
Chris Beall
|
e799c9ffa9
|
tabs converted to two spaces
|
2013-10-10 17:52:57 +00:00 |
Richard Roberts
|
fdced7dbcf
|
Prototype of special comma initializer in unit test only
|
2013-10-10 16:39:15 +00:00 |
Richard Roberts
|
69c7f6d137
|
Restored ClusterTree base class of JunctionTree
|
2013-10-09 17:39:19 +00:00 |
Luca Carlone
|
121e71431a
|
SmartFactorsCreator + initial fix in kitti example
|
2013-10-09 17:13:19 +00:00 |
Alex Cunningham
|
a7e7da49a5
|
Wrapped combine constructors for JacobianFactor and HessianFactor
|
2013-10-08 21:43:14 +00:00 |
Luca Carlone
|
426b4d79cf
|
testSmartProjectionFactor - disabled
|
2013-10-08 14:28:50 +00:00 |
Luca Carlone
|
444df1958a
|
Kitti test not working
|
2013-10-06 18:17:35 +00:00 |
Luca Carlone
|
d049dd38c6
|
Incremental mode in Smart Factors (not working yet with standard ProjectionFactors)
|
2013-10-06 18:17:32 +00:00 |
Frank Dellaert
|
f65740e79e
|
fixed filename in doxygen header
|
2013-10-04 21:28:26 +00:00 |
Zsolt Kira
|
83182b05ca
|
Second method for landmark-first ordering: Provide a constrained ordering with landmarks in group 1 and poses in group 2
|
2013-10-04 14:25:19 +00:00 |
Zsolt Kira
|
cc724134a6
|
Added ability to run both smart and projection factor version (for comparison)
For projection factor version, we can specify a landmark-first ordering
|
2013-10-04 13:47:24 +00:00 |
Richard Roberts
|
43d56b120a
|
From Hmf6_unordered: getISAM2 and variables Reeliminated/Relinearized (d97dcacf186900d12003c399e7fe55e03ca34e3f)
|
2013-10-03 19:52:05 +00:00 |
Richard Roberts
|
0c7b2ec062
|
From Hmf6_unordered: Public isam2 - to be changed (10834edc1698d941be3203e8e6bea0bd4a573443)
|
2013-10-03 19:52:01 +00:00 |
Richard Roberts
|
bf40956592
|
Merged from Hmf6_unordered 970392249a8c50153f24594c9c81acb740cedd06
|
2013-10-03 19:51:56 +00:00 |
Richard Roberts
|
2ad0cd4417
|
Fixed compile error on linux
|
2013-10-03 19:10:13 +00:00 |
Luca Carlone
|
4996db5023
|
Fix in running incremental optimization on Kitti
|
2013-10-03 18:02:25 +00:00 |
Richard Roberts
|
3c68d20ff2
|
Fixed calculate_nnz and added unit test
|
2013-10-03 16:50:20 +00:00 |
Richard Roberts
|
19602bbb0b
|
Fixed warning
|
2013-10-03 16:50:18 +00:00 |
Richard Roberts
|
95ac34aeeb
|
Updating Hessian in-place instead of computing Hessian for each Jacobian
|
2013-10-03 16:50:16 +00:00 |
Richard Roberts
|
4d31dd99f3
|
Hopefully fixed the mex build flags
|
2013-10-03 16:50:14 +00:00 |
Duy-Nguyen Ta
|
c3e3b693e7
|
support special linearize for constraints, which Jacobian could be augmented by Jacobian-from-Hessian parts for Lagrange multipliers in SQP
|
2013-10-02 03:55:11 +00:00 |
Duy-Nguyen Ta
|
63ff1b47c1
|
support augmented matrices in Matrix's vector_scale_in_place, Constrained's Whiten, WhitenInPlace, and unit(...). These are used in constraints SQP solver with Lagrange multipliers.
|
2013-10-02 03:52:08 +00:00 |