You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi @HpWang-whu ,
Thanks for your great work. I have some confusion while reproducing some results from the paper.
In the experiment part, the comparison methods include some rotation averaging methods (e.g. HARA). I wonder how to calculate absolute t for these methods. Does it use the same least squares translation synchronization as the SGHR? If YES, how to determine the weight w for these rotation averaging methods?
For these optimization-based methods, whether the hyperparameters use the original default values or other adjusted values? For example, the RR results of EIGSE3 reported in the paper are 23.2%/60.4% (full/sparse graph with YOHO+3DMatch). However, my reproduced results using the below default EIGSE3 hyperparameters are only 15%/52%.
use_mex=0; % use_mex = false -> do not use mex function to compute weights
use_sparse=1; % use_sparse = true -> use sparse Matlab solvers (eigs/svds)
nmax_eig_irls=50; % nmax = maximum number of iterations of IRLS
thresh_eig_irls=1e-5; % thresh = thrsehold on relative error to check convergence of IRLS
method="top"; % method = 'top' -> compute top eigenvectors
Any help is appreciated!
The text was updated successfully, but these errors were encountered:
For HARA, we simply replace the final rotation synchronization (using L1/2 reweighting function) with SE(3) transformation synchronization.
I certainly made some modifications to the hyperparameters and code of the baselines to achieve their best performances by trial and error. But I am quite sorry that I forgot my final detailed settings of them. :(
1. For HARA, we simply replace the final rotation synchronization (using L1/2 reweighting function) with SE(3) transformation synchronization.
2. I certainly made some modifications to the hyperparameters and code of the baselines to achieve their best performances by trial and error. But I am quite sorry that I forgot my final detailed settings of them. :(
Hi @HpWang-whu ,
Thanks for your great work. I have some confusion while reproducing some results from the paper.
Any help is appreciated!
The text was updated successfully, but these errors were encountered: