Construct a point cloud from the given view data. The data must have a depth channel. All other channels will be stored in t

:return: the alpha channel separated out as a grayscale image

Calculates the noise level (sigma_down) to step down to and the amount of noise to add (sigma_up) when doing an ancestral sampling step.

This is the deprecated API for creating beta schedules. See get_named_beta_schedule() for the new library of schedules.

:return: a tuple of ( the initial output tokens of size [batch_size, data_ctx + latent_ctx, width], an iterator over

:return: a tuple of ( the initial output tokens of size [batch_size, data_ctx + latent_ctx, width], an iterator over

Get a pre-defined beta schedule for the given name. The beta schedule library consists of beta schedules which remain similar in the lim

Constructs the noise schedule of Karras et al. (2022).

:return: coords of shape (width * height, 2)

:return: ([self.height, self.width, 2]).reshape(self.height * self.width, 2) image coordinates

Get output mappings for (h, h_directionless).

Integrate the raw.channels along with other aux_losses and values to produce the final output dictionary containing rendered `channel

:param origin: [batch_size, *shape, 3] :param direction: [batch_size, *shape, 3] :param t0_lower: Optional [batch_size, *shap

depth == 0 is root

For a signed distance field, produce a mesh using marching cubes. :param field: a 3D tensor of field values, where negative values correspon

:param sh_degree: Spherical harmonics degree :param coords_shape: [*shape, 3] :param coords: optional coordinate tensor of coords_shape

Returns an iterator over all the names and meta parameters/buffers. One difference between module.state_dict() is that this preserve

For each point in another set of points, compute the point in this pointcloud which is closest. :param points: an [N x 3] ar

Returns an [N x 3] batch of normal vectors per triangle assuming the right-hand rule.

Sample x_{t-1} from the model at the given timestep. :param model: the model to sample from. :param x: the current tensor at

Generate samples from the model. :param model: the model module. :param shape: the shape of the samples, (N, C, H, W).

Generate samples from the model and yield intermediate samples from each timestep of diffusion. Arguments are the same as p_

Removes the vec2 component from vec1

Given the origin and the direction of a view, return a differentiable projective camera with the given parameters. TODO: We need to supp

Get the distribution q(x_t | x_0). :param x_start: the [N x C x ...] tensor of noiseless inputs. :param t: the number of dif

Returns a [batch x num_views x 1 x size x size] tensor

No op if the image doesn't have an alpha channel. :param: mode: Defaults to "random" but has an option to use a "black" or "white" b

Return an [H x W x 4] RGBA tensor of the rendered image. The pixels are floating points, with alpha in the range [0, 1] and the other col

:param batch: has - rays: [batch_size x ... x 2 x 3] specify the origin and direction of each ray. - radii (optional)

Perform volumetric rendering over the given volume. :param position: [batch_size, *shape, 3] :param direction: [batch_size,

:param: img: image in HWC order :return: currently written for downsampling

Creates a new camera with the same intrinsics and direction as this one, but with resized image dimensions.

Creates a new camera with the same intrinsics and direction as this one, but with resized image dimensions.

:param t0: start time has shape [batch_size, *shape, 1] :param t1: finish time has shape [batch_size, *shape, 1] :param n_sam

:param example: [1, n_points, 3 + n_channels] :return: [1, n_samples, 3 + n_channels]

Sample from the given discrete probability distribution with replacement. The i-th bin is assumed to have mass pmf[i]. :param pmf: [bat

Creates a new camera with the same intrinsics and direction as this one, but with the scene rescaled by the given factor.

Create a list of timesteps to use from an original diffusion process, given the number of timesteps we want to take from equally-sized portio

Returns a [batch x num_views x 1 x size x size] tensor in the range [0, 1].

Returns a [batch x num_views x 3 x size x size] tensor in the range [-1, 1].

Returns a [batch x num_views x 3 x size x size] tensor in the range [-1, 1].