Minor precision fix on unproject

[idilsulo]

Hi! I found a minor issue inside the code sample given for the unproject inside CameraBase.

The provided code block returns True as expected only when the camera is set without any Rotation R or Transformation T.

An example is given below:

torch.set_printoptions(precision=10)

xyz = # 3D points of shape (batch_size, num_points, 3)

R, T = look_at_view_transform(160, 180, 270)
cameras = FoVOrthographicCameras(device=device, R=R, T=T, znear=0.01)

# transform xyz to the camera view coordinates
xyz_cam = cameras.get_world_to_view_transform().transform_points(xyz)
# xyz_cam = cameras.get_full_projection_transform().transform_points(xyz)

# extract the depth of each point as the 3rd coord of xyz_cam
depth = xyz_cam[:, :, 2:]

# project the points xyz to the camera
xy = cameras.transform_points(xyz)[:, :, :2]

# append depth to xy
xy_depth = torch.cat((xy, depth), dim=2)

# unproject to the world coordinates
xyz_unproj_world = cameras.unproject_points(xy_depth, world_coordinates=True, from_ndc=True)

print(xyz)                           # Original set of 3D points
print(xyz_unproj_world)   # Unproject almost identical, except slight precision difference

print(torch.allclose(xyz, xyz_unproj_world)) # Expected: True - Returns: False

I at first thought that R and T were not being considered during unprojection, only to realize afterwards that the print statement above in this case returns False due to precision error.

Changing this line by increasing the absolute threshold atol as also done in the unit tests solves the issue:

pytorch3d/tests/test_cameras.py

Line 619 in d388881

self.assertTrue(torch.allclose(xyz_unproj, matching_xyz, atol=1e-4))

Thanks a lot!

[facebook-github-bot]

Hi @idilsulo!

Thank you for your pull request and welcome to our community.

Action Required

In order to merge any pull request (code, docs, etc.), we require contributors to sign our Contributor License Agreement, and we don't seem to have one on file for you.

Process

In order for us to review and merge your suggested changes, please sign at https://code.facebook.com/cla. If you are contributing on behalf of someone else (eg your employer), the individual CLA may not be sufficient and your employer may need to sign the corporate CLA.

Once the CLA is signed, our tooling will perform checks and validations. Afterwards, the pull request will be tagged with CLA signed. The tagging process may take up to 1 hour after signing. Please give it that time before contacting us about it.

If you have received this in error or have any questions, please contact us at [email protected]. Thanks!

[facebook-github-bot]

Thank you for signing our Contributor License Agreement. We can now accept your code for this (and any) Meta Open Source project. Thanks!

[facebook-github-bot]

Thank you for signing our Contributor License Agreement. We can now accept your code for this (and any) Meta Open Source project. Thanks!

[facebook-github-bot]

Thank you for signing our Contributor License Agreement. We can now accept your code for this (and any) Meta Open Source project. Thanks!

[facebook-github-bot]

@bottler has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator.

[facebook-github-bot]

@idilsulo has updated the pull request. You must reimport the pull request before landing.