Pytorch3d的倒角损失函数Chamfer Distance Loss的用法pytorch3d.loss.chamfer_distance

API文档在这里

源码在这里

之前看到的一个干货满满的Pytorch3D安装指导与简单例子的帖子在这里

 官方tutorials中提到的用法在下面这个代码块里面，前面后面东西都挺多的就把和chamfer_distance相关的摘到了这里

1.  
   from pytorch3d.ops import sample_points_from_meshes
2.  
   from pytorch3d.loss import (
3.  
   chamfer_distance,
4.  
   mesh_edge_loss,
5.  
   mesh_laplacian_smoothing,
6.  
   mesh_normal_consistency,
7.  
   )
8.  
    
9.  
   # We sample 5k points from the surface of each mesh
10.  
    sample_trg = sample_points_from_meshes(trg_mesh, 5000)
11.  
    sample_src = sample_points_from_meshes(new_src_mesh, 5000)
12.  
     
13.  
    # We compare the two sets of pointclouds by computing (a) the chamfer loss
14.  
    loss_chamfer, _ = chamfer_distance(sample_trg, sample_src)

这篇好文章是转载于：学新通技术网