423A35C7/pytorch3d
1
0
Fork 0
mirror of https://github.com/facebookresearch/pytorch3d.git synced 2025-08-02 20:02:49 +08:00
Code Issues Packages Projects Releases Wiki Activity

fix pts scale, save ply

Browse Source 
Summary:
Fix:
* Scaling of point clouds for scalars
* save_ply compatible cat

Reviewed By: nikhilaravi

Differential Revision: D22298609

fbshipit-source-id: abe94a5b64baf325587202d20adfc36912cc1478
This commit is contained in:
Georgia Gkioxari 2020-07-03 10:19:58 -07:00 committed by Facebook GitHub Bot
parent 275ddade66
commit 2f0fd60186
4 changed files with 41 additions and 27 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
pytorch3d/io/ply_io.py
View File

@ -783,7 +783,11 @@ def save_ply(
decimal_places: Number of decimal places for saving. decimal_places: Number of decimal places for saving.
""" """
verts_normals = torch.FloatTensor([]) if verts_normals is None else verts_normals verts_normals = (
torch.tensor([], dtype=torch.float32, device=verts.device)
if verts_normals is None
else verts_normals
)
faces = torch.LongTensor([]) if faces is None else faces faces = torch.LongTensor([]) if faces is None else faces
if len(verts) and not (verts.dim() == 2 and verts.size(1) == 3): if len(verts) and not (verts.dim() == 2 and verts.size(1) == 3):

5
pytorch3d/structures/pointclouds.py
View File

@ -511,7 +511,6 @@ class Pointclouds(object):
Returns: Returns:
1D tensor of indices. 1D tensor of indices.
""" """
self._compute_packed()
if self._padded_to_packed_idx is not None: if self._padded_to_packed_idx is not None:
return self._padded_to_packed_idx return self._padded_to_packed_idx
if self._N == 0: if self._N == 0:
@ -520,7 +519,7 @@ class Pointclouds(object):
self._padded_to_packed_idx = torch.cat( self._padded_to_packed_idx = torch.cat(
[ [
torch.arange(v, dtype=torch.int64, device=self.device) + i * self._P torch.arange(v, dtype=torch.int64, device=self.device) + i * self._P
for (i, v) in enumerate(self._num_points_per_cloud) for (i, v) in enumerate(self.num_points_per_cloud())
], ],
dim=0, dim=0,
) )
@ -797,7 +796,7 @@ class Pointclouds(object):
self. self.
""" """
if not torch.is_tensor(scale): if not torch.is_tensor(scale):
scale = torch.full(len(self), scale) scale = torch.full((len(self),), scale, device=self.device)
new_points_list = [] new_points_list = []
points_list = self.points_list() points_list = self.points_list()
for i, old_points in enumerate(points_list): for i, old_points in enumerate(points_list):

2
tests/test_meshes.py
View File

@ -510,8 +510,8 @@ class TestMeshes(TestCaseMixin, unittest.TestCase):
N = 5 N = 5
for test in ["tensor", "scalar"]: for test in ["tensor", "scalar"]:
for force in (False, True):
mesh = TestMeshes.init_mesh(N, 10, 100) mesh = TestMeshes.init_mesh(N, 10, 100)
for force in [0, 1]:
if force: if force:
# force mesh to have computed attributes # force mesh to have computed attributes
mesh.verts_packed() mesh.verts_packed()

17
tests/test_pointclouds.py
View File

@ -460,7 +460,7 @@ class TestPointclouds(TestCaseMixin, unittest.TestCase):
def test_scale(self): def test_scale(self):
def naive_scale(cloud, scale): def naive_scale(cloud, scale):
if not torch.is_tensor(scale): if not torch.is_tensor(scale):
scale = torch.full(len(cloud), scale) scale = torch.full((len(cloud),), scale, device=cloud.device)
new_points_list = [ new_points_list = [
scale[i] * points.clone() scale[i] * points.clone()
for (i, points) in enumerate(cloud.points_list()) for (i, points) in enumerate(cloud.points_list())
@ -470,18 +470,29 @@ class TestPointclouds(TestCaseMixin, unittest.TestCase):
) )
N = 5 N = 5
clouds = self.init_cloud(N, 100, 10) for test in ["tensor", "scalar"]:
for force in (False, True): for force in (False, True):
clouds = self.init_cloud(N, 100, 10)
if force: if force:
clouds._compute_packed(refresh=True) clouds._compute_packed(refresh=True)
clouds._compute_padded() clouds._compute_padded()
clouds.padded_to_packed_idx() clouds.padded_to_packed_idx()
if test == "tensor":
scales = torch.rand(N) scales = torch.rand(N)
elif test == "scalar":
scales = torch.rand(1)[0].item()
new_clouds_naive = naive_scale(clouds, scales) new_clouds_naive = naive_scale(clouds, scales)
new_clouds = clouds.scale(scales) new_clouds = clouds.scale(scales)
for i in range(N): for i in range(N):
if test == "tensor":
self.assertClose( self.assertClose(
scales[i] * clouds.points_list()[i], new_clouds.points_list()[i] scales[i] * clouds.points_list()[i],
new_clouds.points_list()[i],
)
else:
self.assertClose(
scales * clouds.points_list()[i],
new_clouds.points_list()[i],
) )
self.assertClose( self.assertClose(
clouds.normals_list()[i], new_clouds_naive.normals_list()[i] clouds.normals_list()[i], new_clouds_naive.normals_list()[i]