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
Implemented high-performance CUDA kernel grid::create_voxel_cloud_corners_kernel in conquer3d/csrc/data_structure/grid.cu and grid.h.
Added PyBind11 C++ backend function create_voxel_cloud_from_tmesh with row-wise at::unique_dim corner deduplication in grid.cpp and grid.py.
Implemented high-level tmesh2voxelcloud conversion API in conquer3d/conversion/tmesh.py for constructing non-rigid, vertex-centered overlapping 3D voxel clouds.
Removed obsolete iso parameter from both tmesh2sparse and tmesh2voxelcloud APIs.
Stanford PLY Support (conquer3d/io/ply.py): Added high-speed loaders and writers read_ply, write_ply, and write_voxel_ply.
Wavefront OBJ Quad Support (conquer3d/io/obj.py): Added write_quad_obj and write_voxel_obj routines to export 4-vertex quad faces and 6-quad per voxel 3D cell representations.
Exported all PLY and OBJ functions in conquer3d/io/__init__.py.