Geman-Mcclure robost loss

[Minh Vo]

Hello,

I wonder what is the best way to implement the Geman-Mcclure (GM) loss (i.e., /rho = (s^2/(s^2+c)). Since the Ceres assumes the robust loss is applied on the squared norm, one way is to take the square root of the GM loss and feed that to the null loss. However, this approach is not clean and it also requires redundant computation.

Thanks,

Minh

[Sameer Agarwal]

Minh,

Why not just implement a loss function that does Geman-Mcclure and use it. Where is the square root here?

Sameer

--
You received this message because you are subscribed to the Google Groups "Ceres Solver" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ceres-solver...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/ceres-solver/32752ae5-6b12-4704-972a-3f109bf8d77e%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Minh Vo]

Got it.... I was being so silly. Thanks!