For each object in AI2-THOR metadata, update with manually defined object classes to be tracked in custom properties

Return Euclidean distance between two agents in the sim.

Add values to custom properties that are lists

Return available AI2-THOR floor plans, restricting to world_type if provided :param world_type: One of "Kitchen", "Bedroom", "Bathroo

Obtain the nearest object to the commander OR driver matching the given search string. :param query: the search string to check again

Identify an object around relative coordinate (x, y) in the egocentric frame that satisfies required affordances :param x: Relative c

Construct the overall mAP metric. This will compute the following metrics: - Box + Noun mAP - Box + Noun + Verb mAP

Construct the overall mAP metric. This will compute the following metrics: - Box + Noun mAP - Box + Noun + Verb mAP

Initialize simulator for a new episode

Compute minimum distance between two sets of points.

Create a grayscale version of the original image. The colors in masked area, if given, will be kept.

Create a grayscale version of the original image. The colors in masked area, if given, will be kept.

Find proper places to draw text given a binary mask.

Find proper places to draw text given a binary mask.

Actually step the environment with the provided action. Return (None if STOP, action_str, screenshot_b64, action_output).

Return an activation function given a string

Return an activation function given a string

Return instance ID

Convert a screenshot (numpy array) to base64.

Divide a by b avoiding a DivideByZero exception Inputs: a, b: either vectors or scalars Outputs: eith

Divide a by b avoiding a DivideByZero exception Inputs: a, b: either vectors or scalars Outputs: eith

Computes the precision@k for the specified values of k

Add predictions and labels of a single image and matches predictions to ground truth boxes :param predictions: dictionary of predicti

salem: dilation structure normally used to dilate the map for path planning

Send keystrokes to the focused element.

max_openable_per_cat: how many max of each category to get observations for

Perform backwarding, gathering the gradients across different process/ GPU group.

corners: (8,3) no assumption on axis direction

Perform center crop on input images. Args: size (int): size of the cropped height and width. image (array): the image to perf

Perform center crop on input images. Args: size (int): size of the cropped height and width. image (array): the image to perf

rgb: single rgb image NOTE: currently this only handles a single RGB

further clean accesibility tree

Compare a value (int or float) against an inequality string. Args: - value (int/float): The value to be compared. -

Compute a matrix of intersection over union values for two lists of bounding boxes using broadcasting :param preds: matrix of predicted bound