Create a temporary table and index it to speed up repeated SQL quesries. The table is an INNER JOIN of CUPTI_ACTIVITY_KIND_MARKER with itself.

Demangle a C++ string

detach_hidden()

'semseg-format' vs. 'labels'

Get for all splits, at once. Args: - required_class_names - highlight_classname: class to highlight in pink - condition - folder_pref

Write out all requested COCO masks to disk. Get for all splits, at once. If person-car combinations are desired, this tuple may be specified

Convert PASCAL Context annotations from .mat files to .png Args: - pcontext_dst_dir: string represent absolute path to PASCAL Context dest

Given process id (pid) and thread id (tid), return the object id. object id = pid (little endian 4 bytes) + tid (little endian 8 bytes)

Calculate accuracy. Args: - save_vis: whether to save visualizations on predictions vs. ground truth

Calculate accuracy, although use both relabeld and unrelabeled data to get a number that is relatable to unrelabeled baselines TODO:

id is excluded id take in array of iou/acc., return non-excluded values

Args: - test_loader: Returns: - None

input_file is a path to a video file. Read frames from an RGB video file, and write overlaid predictions into a new video file. Args:

Args: - num_processes: number of workers to exploit Returns: - None

For any given dataset, compare relabeled classes (in universal taxonomy) with universal classes that correspond with original classes. We populat

Frees all allocated memory chunks */

Given the correlation id, get CPU start, end, thread id, process id. The information can be in the runtime table or the driver table.

Get GPU kernel info

This function first finds all NVTX markers encapsulating a runtime / driver kernel launch. It then splits the markers into many lists. layer

Get the profile start time

Returns a shorter kernel name

See use here in Pytorch ImageNet script: https://github.com/pytorch/examples/blob/master/imagenet/main.py#L197 Returns: -

Get COCO Panoptic instance annotations from .pngs Return a list of filepaths to the instance ID images Args: - split: string representi

Args: - fpath: string representing file path - n: integer representing last number of filepath elements to keep Returns: -

gets params of reparameterization based on known attribute names

TODO: add to library to avoid replication.

TODO: add to library to avoid replication.

Args: - dname: name of dataset to apply re-labeling to Returns: - DatasetRewritingTask to complete

Args: - dname: name of dataset to apply re-labeling to Returns: - DatasetRewritingTask to complete

e.g. leftImg8bit/train/0/022141_leftImg8bit.png e.g. leftImg8bit/train/aachen/aachen_000000_000019_leftImg8bit.png Args: - seq_id -

Args: - split: string representing training, validation, or testing split of the data - fname_stem: Returns: - rgb_img: color image.

Args: - None Returns: - split_dict - num_imgs_per_split_dict - num_scans_per_split_dict