Open gtangg11 opened 2 months ago
Thank you for your correction! I've fixed the bug caused by the igl version change. Regarding the np.vstack issue, I first reshaped the input tensor into a 2D array and managed the batch splitting, so I think it should be fine. Could you clarify how this might cause a batching issue?
in extract_meshes, due to igl (version you're using doesn't build on certain systems) update, we should have
v, _, _, f = igl.remove_duplicate_vertices(np.array(octree.mesh_v), tri_faces, 1e-7)
and also np.vstack should be np.concatenate to avoid batching issues.
Putting these fixes in, I was able to run the code. Otherwise it crashes. Please address these changes. Thank you!