Deprecate so3_exponential_map

Summary: Deprecate the `so3_exponential_map()` function in favor of its alias `so3_exp_map()`: this aligns with the naming of `so3_log_map()` and the recently introduced `se3_exp_map()` / `se3_log_map()` pair.

Reviewed By: bottler

Differential Revision: D29329966

fbshipit-source-id: b6f60b9e86b2995f70b1fbeb16f9feea05c55de9

tests/test_camera_conversions.py

@@ -12,7 +12,7 @@ import numpy as np
 import torch
 from common_testing import TestCaseMixin, get_tests_dir
 from pytorch3d.ops import eyes
-from pytorch3d.transforms import so3_exponential_map, so3_log_map
+from pytorch3d.transforms import so3_exp_map, so3_log_map
 from pytorch3d.utils import (
     cameras_from_opencv_projection,
     opencv_from_cameras_projection,
@@ -33,7 +33,7 @@ def cv2_project_points(pts, rvec, tvec, camera_matrix):
     """
     Reproduces the `cv2.projectPoints` function from OpenCV using PyTorch.
     """
-    R = so3_exponential_map(rvec)
+    R = so3_exp_map(rvec)
     pts_proj_3d = (
         camera_matrix.bmm(R.bmm(pts.permute(0, 2, 1)) + tvec[:, :, None])
     ).permute(0, 2, 1)

tests/test_cameras.py

@@ -53,7 +53,7 @@ from pytorch3d.renderer.cameras import (
 )
 from pytorch3d.transforms import Transform3d
 from pytorch3d.transforms.rotation_conversions import random_rotations
-from pytorch3d.transforms.so3 import so3_exponential_map
+from pytorch3d.transforms.so3 import so3_exp_map
 
 
 # Naive function adapted from SoftRasterizer for test purposes.
@@ -145,7 +145,7 @@ def init_random_cameras(
     T = torch.randn(batch_size, 3) * 0.03
     if not random_z:
         T[:, 2] = 4
-    R = so3_exponential_map(torch.randn(batch_size, 3) * 3.0)
+    R = so3_exp_map(torch.randn(batch_size, 3) * 3.0)
     cam_params = {"R": R, "T": T}
     if cam_type in (OpenGLPerspectiveCameras, OpenGLOrthographicCameras):
         cam_params["znear"] = torch.rand(batch_size) * 10 + 0.1
@@ -509,7 +509,7 @@ class TestCamerasCommon(TestCaseMixin, unittest.TestCase):
     def init_equiv_cameras_ndc_screen(cam_type: CamerasBase, batch_size: int):
         T = torch.randn(batch_size, 3) * 0.03
         T[:, 2] = 4
-        R = so3_exponential_map(torch.randn(batch_size, 3) * 3.0)
+        R = so3_exp_map(torch.randn(batch_size, 3) * 3.0)
         screen_cam_params = {"R": R, "T": T}
         ndc_cam_params = {"R": R, "T": T}
         if cam_type in (OrthographicCameras, PerspectiveCameras):

tests/test_cameras_alignment.py

@@ -17,7 +17,7 @@ from pytorch3d.renderer.cameras import (
     SfMPerspectiveCameras,
 )
 from pytorch3d.transforms.rotation_conversions import random_rotations
-from pytorch3d.transforms.so3 import so3_exponential_map, so3_relative_angle
+from pytorch3d.transforms.so3 import so3_exp_map, so3_relative_angle
 from test_cameras import init_random_cameras
 
 
@@ -95,9 +95,7 @@ class TestCamerasAlignment(TestCaseMixin, unittest.TestCase):
         ) * s_align_gt
 
         if add_noise != 0.0:
-            R_new = torch.bmm(
-                R_new, so3_exponential_map(torch.randn_like(T_new) * add_noise)
-            )
+            R_new = torch.bmm(R_new, so3_exp_map(torch.randn_like(T_new) * add_noise))
             T_new += torch.randn_like(T_new) * add_noise
 
         # create new cameras from R_new and T_new

tests/test_points_to_volumes.py

@@ -15,7 +15,7 @@ from pytorch3d.ops.sample_points_from_meshes import sample_points_from_meshes
 from pytorch3d.structures.meshes import Meshes
 from pytorch3d.structures.pointclouds import Pointclouds
 from pytorch3d.structures.volumes import Volumes
-from pytorch3d.transforms.so3 import so3_exponential_map
+from pytorch3d.transforms.so3 import so3_exp_map
 
 
 DEBUG = False
@@ -138,7 +138,7 @@ def init_uniform_y_rotations(batch_size: int = 10):
     angles = torch.linspace(0, 2.0 * np.pi, batch_size + 1, device=device)
     angles = angles[:batch_size]
     log_rots = axis[None, :] * angles[:, None]
-    R = so3_exponential_map(log_rots)
+    R = so3_exp_map(log_rots)
     return R
 
 

tests/test_r2n2.py

@@ -29,7 +29,7 @@ from pytorch3d.renderer import (
 )
 from pytorch3d.renderer.cameras import get_world_to_view_transform
 from pytorch3d.transforms import Transform3d
-from pytorch3d.transforms.so3 import so3_exponential_map
+from pytorch3d.transforms.so3 import so3_exp_map
 from torch.utils.data import DataLoader
 
 
@@ -316,7 +316,7 @@ class TestR2N2(TestCaseMixin, unittest.TestCase):
         """
         # Test get_world_to_view_transform.
         T = torch.randn(10, 3)
-        R = so3_exponential_map(torch.randn(10, 3) * 3.0)
+        R = so3_exp_map(torch.randn(10, 3) * 3.0)
         RT = get_world_to_view_transform(R=R, T=T)
         cam = BlenderCamera(R=R, T=T)
         RT_class = cam.get_world_to_view_transform()

tests/test_transforms.py

@@ -10,7 +10,7 @@ import unittest
 
 import torch
 from common_testing import TestCaseMixin
-from pytorch3d.transforms.so3 import so3_exponential_map
+from pytorch3d.transforms.so3 import so3_exp_map
 from pytorch3d.transforms.transform3d import (
     Rotate,
     RotateAxisAngle,
@@ -146,7 +146,7 @@ class TestTransform(TestCaseMixin, unittest.TestCase):
         self.assertTrue(torch.allclose(normals_out, normals_out_expected))
 
     def test_rotate(self):
-        R = so3_exponential_map(torch.randn((1, 3)))
+        R = so3_exp_map(torch.randn((1, 3)))
         t = Transform3d().rotate(R)
         points = torch.tensor([[1.0, 0.0, 0.0], [0.0, 1.0, 0.0], [0.5, 0.5, 0.0]]).view(
             1, 3, 3
@@ -273,7 +273,7 @@ class TestTransform(TestCaseMixin, unittest.TestCase):
                     )
                 elif choice <= 2.0 / 3.0:
                     t_ = Rotate(
-                        so3_exponential_map(
+                        so3_exp_map(
                             torch.randn(
                                 (batch_size, 3), dtype=torch.float32, device=device
                             )
@@ -894,7 +894,7 @@ class TestRotate(unittest.TestCase):
     def test_inverse(self, batch_size=5):
         device = torch.device("cuda:0")
         log_rot = torch.randn((batch_size, 3), dtype=torch.float32, device=device)
-        R = so3_exponential_map(log_rot)
+        R = so3_exp_map(log_rot)
         t = Rotate(R)
         im = t.inverse()._matrix
         im_2 = t._matrix.inverse()