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

Implicit function docfix

Browse Source 
Summary: Fixes implicit function doc.

Reviewed By: theschnitz, nikhilaravi

Differential Revision: D26870946

fbshipit-source-id: 5d03ebbc284153c41b9d6695b28c8b4e11bc0a5c
This commit is contained in:
David Novotny 2021-04-08 07:43:15 -07:00 committed by Facebook GitHub Bot
parent cc08c6b288
commit 7c0d307142
1 changed files with 20 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

26
pytorch3d/renderer/implicit/renderer.py
View File

@ -32,23 +32,29 @@ class ImplicitRenderer(torch.nn.Module):
VOLUMETRIC_FUNCTION VOLUMETRIC_FUNCTION
The `forward` function of the renderer accepts as input the rendering cameras as well The `forward` function of the renderer accepts as input the rendering cameras
as the `volumetric_function` `Callable`, which defines a field of opacity as well as the `volumetric_function` `Callable`, which defines a field of opacity
and feature vectors over the 3D domain of the scene. and feature vectors over the 3D domain of the scene.
A standard `volumetric_function` has the following signature: A standard `volumetric_function` has the following signature:
``` ```
def volumetric_function(ray_bundle: RayBundle) -> Tuple[torch.Tensor, torch.Tensor] def volumetric_function(
ray_bundle: RayBundle,
**kwargs,
) -> Tuple[torch.Tensor, torch.Tensor]
``` ```
With the following arguments: With the following arguments:
`ray_bundle`: A RayBundle object containing the following variables: `ray_bundle`: A RayBundle object containing the following variables:
`rays_origins`: A tensor of shape `(minibatch, ..., 3)` denoting `origins`: A tensor of shape `(minibatch, ..., 3)` denoting
the origins of the rendering rays. the origins of the rendering rays.
`rays_directions`: A tensor of shape `(minibatch, ..., 3)` `directions`: A tensor of shape `(minibatch, ..., 3)`
containing the direction vectors of rendering rays. containing the direction vectors of rendering rays.
`rays_lengths`: A tensor of shape `lengths`: A tensor of shape
`(minibatch, ..., num_points_per_ray)`containing the `(minibatch, ..., num_points_per_ray)`containing the
lengths at which the ray points are sampled. lengths at which the ray points are sampled.
`xys`: A tensor of shape
`(minibatch, ..., 2)` containing the
xy locations of each ray's pixel in the screen space.
Calling `volumetric_function` then returns the following: Calling `volumetric_function` then returns the following:
`rays_densities`: A tensor of shape `rays_densities`: A tensor of shape
`(minibatch, ..., num_points_per_ray, opacity_dim)` containing `(minibatch, ..., num_points_per_ray, opacity_dim)` containing
@ -57,12 +63,20 @@ class ImplicitRenderer(torch.nn.Module):
`(minibatch, ..., num_points_per_ray, feature_dim)` containing `(minibatch, ..., num_points_per_ray, feature_dim)` containing
the an feature vector for each ray point. the an feature vector for each ray point.
Note that, in order to increase flexibility of the API, we allow multiple
other arguments to enter the volumentric function via additional
(optional) keyword arguments `**kwargs`.
A typical use-case is passing a `CamerasBase` object as an additional
keyword argument, which can allow the volumetric function to adjust its
outputs based on the directions of the projection rays.
Example: Example:
A simple volumetric function of a 0-centered A simple volumetric function of a 0-centered
RGB sphere with a unit diameter is defined as follows: RGB sphere with a unit diameter is defined as follows:
``` ```
def volumetric_function( def volumetric_function(
ray_bundle: RayBundle, ray_bundle: RayBundle,
**kwargs,
) -> Tuple[torch.Tensor, torch.Tensor]: ) -> Tuple[torch.Tensor, torch.Tensor]:
# first convert the ray origins, directions and lengths # first convert the ray origins, directions and lengths