6b890cec0e
added updateNoiseModels functionality
2014-09-10 00:24:32 +03:00
70956bb447
minor fixes for clang in C++11 mode on Mac
2014-09-08 15:52:07 -04:00
5bc4810dcb
Correcting bug fixes for Visual Studio. See bb issues #115,116,118 for more information
2014-09-08 15:19:54 -04:00
e3ec28ebeb
added debug info for gradient code
2014-09-06 18:27:07 -04:00
41bb99b48a
fixed gradient
2014-09-06 10:57:22 -04:00
aacd3484c7
bug fix: different pose covariance structure between g2o (t,R) and gtsam (R,t)
2014-09-04 10:07:28 -04:00
8444680e6e
fixed unit test
2014-09-03 20:20:06 -04:00
08a6bff803
fixed issues in dataset and added comments
2014-09-03 20:18:46 -04:00
544e56f634
added examples
2014-09-03 15:40:57 -04:00
316ce41a22
removed comments
2014-09-03 15:31:30 -04:00
b19e758f5a
no test failures
2014-09-03 15:30:41 -04:00
103d2a8ae9
working unit tests
2014-09-03 15:23:35 -04:00
35d5b56b65
still debugging
2014-09-02 20:56:36 -04:00
8384c37179
fixed 1 unit test
2014-09-02 20:17:31 -04:00
366101176c
added initial failing unit test
2014-09-02 18:50:23 -04:00
15abbc484b
setting up the gradient method
2014-09-02 18:38:12 -04:00
16bb90387c
added example optimizing rotations only from g2o file
2014-08-31 13:03:31 -04:00
f6adeb8fff
working example
2014-08-25 16:36:58 -04:00
e605c2dbc5
added initialization example
2014-08-25 16:01:53 -04:00
330fbd730e
fix failing unit test (mac)
2014-08-22 10:40:16 -04:00
3bb9b38c1f
add "between" so Cal3_S2 can be used with BetweenFactor
2014-08-20 13:05:14 -04:00
889bbea8d4
finished and tested initializer
2014-08-19 21:09:51 -04:00
fdb31dbd8d
Merge branch 'develop'
2014-08-19 20:51:16 -04:00
f5cc247b1c
added pose initializer
2014-08-19 20:50:48 -04:00
bc205cf6a4
extended g2o parser to nondiagonal noise
2014-08-19 20:10:42 -04:00
881514a45f
added 3D version of writeG2o with unit test
2014-08-18 21:14:48 -04:00
358c978f00
added 3D version of readG2o with unit test
2014-08-18 20:40:52 -04:00
c0f880d52b
working version with pose
2014-08-17 20:50:18 -04:00
4cd4023ef8
included normalization
2014-08-17 20:34:43 -04:00
e031ba036d
first working version with (unnormalized) rotations only
2014-08-17 20:24:53 -04:00
2f2a40a737
initial implementation of chordal relaxation
2014-08-17 20:02:46 -04:00
b58d377a97
Small rename for clarity.
2014-08-15 17:34:22 +02:00
456c8eb6ed
Partial fix for Boost 1.56 with Visual Studio 2013. More errors remain.
2014-08-14 00:57:41 -04:00
7f3aaec014
Remove unimportant part of test: different result with boost 1.56, and it doesn't matter
2014-08-14 00:33:57 -04:00
7a658e0347
relaxed tolerance a tiny bit to make test pass with Boost 1.56
2014-08-14 00:25:49 -04:00
24f1da0e01
Fixes for latest boost 1.56 on Linux
2014-08-14 00:15:15 -04:00
2634b2e890
Finish concurrent-calibration
2014-08-12 16:00:20 -04:00
50c8313233
Rename TransformProjectionFactor to ProjectionFactorPPP
2014-08-12 15:55:28 -04:00
2c3f6e563d
Renamed TransformCalProjectionFactor to ProjectionFactorPPPC to follow new naming convention
2014-08-12 15:29:05 -04:00
f8126dbf78
fix .cproject after rebasing
2014-08-06 22:41:35 -04:00
af54d39b13
ignore build folder
2014-08-06 11:10:36 -04:00
779a21c2c4
A small example showing that Hessian matrices in Lie groups might not be symmetric
2014-08-06 11:10:36 -04:00
3778e3c928
constructors for JacobianFactor up to 6-ary for testing in matlab
2014-08-06 11:10:36 -04:00
ecb896ef03
move Adjoint to cpp and enable EXP_MAP
2014-08-06 11:10:36 -04:00
c6d541741b
more comments and debug info
2014-08-06 11:10:36 -04:00
47ed9f3687
matlab code to reproduce test results
2014-08-06 11:10:36 -04:00
4037d1ec1a
wrap QPSolver
2014-08-06 11:10:35 -04:00
cc2ba1792d
add detailed comments for the new EliminatePreferCholesky
2014-08-06 11:10:35 -04:00
a7c0d10862
disable two warning options in METIS which are not understood by my clang compiler.
2014-08-06 11:10:35 -04:00
2895b0515c
Support non positive definite Hessian factors while doing EliminatePreferCholesky with some constrained factors.
...
Currently, when eliminating a constrained variable, EliminatePreferCholesky converts every other factors to JacobianFactor before doing the special QR factorization for constrained variables. Unfortunately, after a constrained nonlinear graph is linearized, new hessian factors from constraints, multiplied with the dual variable (-lambda*\hessian{c} terms in the Lagrangian objective function), might become negative definite, thus cannot be converted to JacobianFactors.
Following EliminateCholesky, this version of EliminatePreferCholesky for constrained var gathers all unconstrained factors into a big joint HessianFactor before converting it into a JacobianFactor to be eliminiated by QR together with the other constrained factors.
Of course, this might not solve the non-positive-definite problem entirely, because (1) the original hessian factors might be non-positive definite and (2) large strange value of lambdas might cause the joint factor non-positive definite [is this true?]. But at least, this will help in typical cases.
2014-08-06 11:09:25 -04:00
Vadim Indelman
cbeall3
Andrew Melim
Luca
dellaert
thduynguyen