Closed ihmc3jn09hk closed 3 years ago
Hi Robin, thank you for sharing the work. From the main-page, there suggests
mesh1 = TriMesh('data/cat-00.off') mesh2 = TriMesh('data/lion-00.off') ... model.fit(descr_mu=descr_mu, lap_mu=lap_mu, descr_comm_mu=descr_comm_mu, verbose=True) # One can access the functional map FM and vertex to vertex mapping p2p FM = model.FM p2p = model.p2p ... print (p2p.shape) # shows (5000,) not (7207,) or (7207, 5000)
I am wondering what the
p2p
is acting on ? Is it a correspondence 'matrix' from mesh1->mesh2 ? Since mesh1 has 7202 vertices, I know there will be some surjective map from mesh1->mesh2. However, when checking the dimension of the returned p2p, the dimension is neither (7207,) or (7207, 5000). How should I get the correspondence correctly?
I think I found the mistake. I have miss interpreted the p2p as mesh1->mesh2. It actually is mesh2->mesh1.
Hi,
Yes the p2p map is going to the direction opposite to the functional map.
This is actually the case in the whole code for clarity, but I should mention it somewhere in the main page.
Hi Robin, thank you for sharing the work. From the main-page, there suggests
I am wondering what the
p2p
is acting on ? Is it a correspondence 'matrix' from mesh1->mesh2 ? Since mesh1 has 7202 vertices, I know there will be some surjective map from mesh1->mesh2. However, when checking the dimension of the returned p2p, the dimension is neither (7207,) or (7207, 5000). How should I get the correspondence correctly?