saberbud/ceres-solver
1
0
Fork 0
mirror of https://github.com/ceres-solver/ceres-solver.git synced 2026-08-30 00:50:37 +08:00
Code Issues Packages Projects Releases Wiki Activity

Speed up the application of robust loss functions.

Browse Source 
Since we added special handling for the case for rho[2] < 0,
the bulk of CorrectJacobian is pointless in the common case.
So add a simple one dimensional loop which rescales the Jacobian.
This speeds up this method immensely.

The robustification of a Jacobian gets speeded up by > 50%.

Change-Id: I97c4e897ccbb5521c053e1fb931c5d0d32f542c7
This commit is contained in:
Sameer Agarwal
2013-11-08 08:04:44 -08:00
parent 58792dc8ee
commit 8c155d51fa
2 changed files with 15 additions and 9 deletions
Download Patch File Download Diff File Expand all files Collapse all files
include/ceres/jet.h
+2 -2
View File
@@ -106,8 +106,8 @@
// Jet<double, 2> y(1); // Pick the 1st dual number for y.
// Jet<double, 2> z = f(x, y);
//
// LG << "df/dx = " << z.a[0]
// << "df/dy = " << z.a[1];
// LOG(INFO) << "df/dx = " << z.a[0]
// << "df/dy = " << z.a[1];
//
// Most users should not use Jet objects directly; a wrapper around Jet objects,
// which makes computing the derivative, gradient, or jacobian of templated