Convert the LVIS class name into a query text for MDETR

This functions accepts a dict from the 'dataset_name' to the list of datapoints we have for this dataset. It splits the data points based on wheth

Do the heavy lifting on the given subset (eg 'train')

Do the heavy lifting on the given split (eg 'train')

Do the heavy lifting on the given split (eg 'train')

Accepts a list of bounding boxes in format (xmin,ymin,xmax,ymax) Returns the list of boxes in coco format (xmin,ymin, width, height)

Accepts a list of bounding boxes in format (xmin,ymin,xmax,ymax) Returns the list of boxes in coco format (xmin,ymin, width, height)

construct a map such that positive_map[i,j] = True iff box i is associated to token j

This functions accepts pre-processed region descriptions for a given image, and removes regions that are redundant. Two regions are deemed redunda

Compute the DICE loss, similar to generalized IOU for masks Args: inputs: A float tensor of arbitrary shape. The pred

Download images from mscoco.org server. Args: save_dir (str): dir to save downloaded images img_ids (int array): img i

Run per image evaluation on given images and store results (a list of dict) in self.evalImgs :return: None

Perform evaluation for single category and image.

Return the template that corresponds to the given question

Return the template that corresponds to the given refexp

Parses the xml files in the Flickr30K Entities dataset input: filename - full file path to the annotations file to parse output:

Get all category ids. Returns: ids (int array): integer array of category ids

Find clusters of highly overlapping boxes Parameters: - all_boxes: a list of boxes in [center_x, center_y, w, h] format - iou_thre

Get all img ids. Returns: ids (int array): integer array of image ids

get a batch with one image and all refer data on that image

Returns: The rank of the current process within the global process group.

Parses a sentence file from the Flickr30K Entities dataset input: filename - full file path to the sentence file to parse output:

Parses a sentence file from the Flickr30K Entities dataset input: filename - full file path to the sentence file to parse output:

Return true if the current process is the main one

Load categories with the specified ids. If ids=None load all categories. Args: ids (int array): integer array of category

Loads the CLEVR templates from disk