Visualize fake normals of `depth` (optionally scaled to be isotropic).

Compute weighted FLIP histogram :param flip_error_map: matrix (with HxW layout) containing per-pixel FLIP values in the [0,1] range :param direct

Add a branch to a node. Split the node if necessary. Returns 0 if node not split. Old node updated. Returns 1 if node split, sets *new_node to addre

Allocate space for a node in the list used in DeletRect to store Nodes that are too empty.

Use one of the methods to calculate retangle volume

Method #0 for choosing a partition: As the seeds for the two groups, pick the two rects that would waste the most area if covered by a single rectangl

Put a branch in one of the groups.

Combine two rectangles into larger one containing both

Disconnect a dependent node. Caller must return (or stop using iteration index) after this as count has changed

Find the next data element in the tree (For internal use only)

finds position of least significant unset bit */

Get the bounds for this node

Load branch buffer with branches from full node plus the extra branch.

Get 'first' for iteration

Get Next for iteration

Initialize a PartitionVars structure.

Insert a data rectangle into an index structure. InsertRect provides for splitting the root; returns 1 if root was split, 0 if it was not. The level a

Inserts a new data rectangle into the index structure. Recursively descends tree, propagates splits back up. Returns 0 if node was not split. Old nod

Is iterator pointing to valid data

Is iterator invalid

Is iterator NULL, or at end?

Constructor: takes a const ref to the matrix object with the data points

Deleted copy constructor*/

* KDTree constructor * * Refer to docs in README.md or online in * https://github.com/jlblancoc/nanoflann * * The KD-Tree point dimensi

* KDTree constructor * * Refer to docs in README.md or online in * https://github.com/jlblancoc/nanoflann * * The KD-Tree point dimensi

Copy branches from the buffer into two nodes according to the partition.

Find the smallest rectangle that includes all rectangles in branches of a node.

Decide whether two rectangles overlap.

Pick a branch. Pick the one that will need the smallest increase in area to accomodate the new rectangle. This will result in the least total area f

Default constructor. Initializes a new pool. */

Add a node to the reinsertion list. All its branches will later be reinserted into the index structure.

The exact volume of the bounding sphere for the given Rect

Calculate the n-dimensional volume of a rectangle

Delete a data rectangle from an index structure. Pass in a pointer to a Rect, the tid of the record, ptr to ptr to root node. Returns 1 if record not

Delete a rectangle from non-root part of an index structure. Called by RemoveRect. Descends tree recursively, merges branches on the way back up. Ret

Note: The first argument (checks_IGNORED_) is ignored, but kept for * compatibility with the FLANN interface */

Split a node. Divides the nodes branches and the extra one between two nodes. Old node is one of the new ones, and one really new one is created. Trie

Nerf Model. Args: rng_0: jnp.ndarray, random number generator for coarse model sampling. rng_1: jnp.ndarray, random number generator

Args: x: jnp.ndarray, [batch, 3] Returns:

Evaluate the MLP. Args: x: jnp.ndarray(float32), [batch, num_samples, feature], points. condition: jnp.ndarray(float32), [batch, feat

Evaluate the MLP. Args: x: jnp.ndarray(float32), [batch, num_samples, feature], points. condition: jnp.ndarray(float32), [batch, feat

r""" class for ms-ssim Args: win_size: (int, optional): the size of gauss kernel win_sigma: (float, optional): sigma o

Get the next training batch or test example. Returns: batch: dict, has "pixels" and "rays".

Generating rays for all images.

Generate normalized device coordinate rays for llff.

Load images from disk.

Load images from disk.

Load images from disk.

Sample next test example.

Sample next test example.

Sample next training batch.