Methods for encoding point cloud feature

A computer device, comprising:

• a memory;
• and
• one or more processors;
• wherein computer-readable instructions are stored in the memory, when executed by the one or more processors, causing the one or more processors to perform a method comprising:
• obtaining an original point cloud;
• generating point-wise hybrid scale voxel features by conducting hybrid scale voxelization on the original point cloud;
• generating voxel-wise attention features using spatial features and voxel features of points in a voxel at each voxel scale;
• aggregating the point-wise hybrid scale voxel features and the voxel-wise attention features to obtain voxel-wise projection scale information;
• mapping the voxel-wise projection scale information to pseudo image features at projection scales;
• and
• encoding point-wise scale voxel features at one of the voxel scales and scale voxel features of the same point at other voxel scales to generate the point-wise hybrid scale voxel features.

One or more non-transitory computer-readable storage medium storing computer-readable instructions, when executed by the one or more processors, causing the one or more processors to perform a method comprising:

• obtaining an original point cloud;
• generating point-wise hybrid scale voxel features by conducting hybrid scale voxelization on the original point cloud;
• generating voxel-wise attention features using spatial features and voxel features of points in a voxel at each voxel scale;
• aggregating the point-wise hybrid scale voxel features and the voxel-wise attention features to obtain voxel-wise projection scale information;
• mapping the voxel-wise projection scale information to pseudo image features at projection scales;
• and
• encoding point-wise scale voxel features at one of the voxel scales and scale voxel features of the same point at other voxel scales to generate the point-wise hybrid scale voxel features.