CIntFromPy */

CIntToPy */

CIntToPy */

SliceObject */

RaiseArgTupleInvalid */

ExtTypeTest */

CodeObjectCache */

CodeObjectCache */

Compute object category assignment for each box associated with each roidb entry.

Checks that `value_a`, which is intended to replace `value_b` is of the right type. The type is correct if it matches exactly or is one of a few

Decodes a raw config value (e.g., from a yaml config files or command line argument) into a Python object.

Bounding-box regression targets are stored in a compact form in the roidb. This function expands those targets into the 4-of-4*K representati

Finds proposals that are inside crowd regions and marks them as overlap = -1 with each ground-truth rois, which means they will be excluded fr

Convert an image and RoIs within that image into network inputs.

Merge config dictionary a into config dictionary b, clobbering the options in b whenever they are also specified in a.

Add proposal boxes to each roidb entry.

Given a vector of widths (ws) and heights (hs) around a center (x_ctr, y_ctr), output a set of anchors (windows).

Check if the dataset uses the Cityscapes dataset evaluator.

Check if the dataset uses the general json dataset evaluator.

Return width, height, x center, and y center for an anchor (window).

Check that cache file has the correct hash.

Given the ratio_list, we want to make the RATIO same for each minibatch on each GPU. Note: this only work for 1) cfg.TRAIN.MAX_SIZE is ignored dur

Load and concatenate roidbs for one or more datasets, along with optional object proposals. The roidb entries are then prepared for use in trainin

r""" Gathers variables from different GPUs on a specified device (-1 means the CPU).

Calculate target resize scale

Process the outputs of model for testing Args: model: the network module im_data: Pytorch variable. Input batch to the model. im

Computes bbox detections at the given width-relative aspect ratio. Returns predictions in the original image space.

Computes bbox detections at the given scale. Returns predictions in the original image space.

Detects keypoints at the given width-relative aspect ratio.

Computes keypoint predictions at the given scale.

Computes mask detections at the given width-relative aspect ratio.

Computes masks at the given scale.

Log training statistics to terminal

Construct weight mapping relation for a residual stage with `num_blocks` of residual blocks given the stage id: `res_id`

r""" Slices variables into approximately equal chunks and distributes them across given GPUs. Duplicates references to objects that are no

r"""Scatter with support for kwargs dictionary

Share res5 block with box head on training

Log the tracked statistics to tensorboard

Log the tracked statistics.

proto*/

ResNet "conv5" / "stage5" head for predicting masks.

Update tracked iteration statistics for the case of iter_size > 1

BufferFormatCheck */

BufferFormatCheck */

CLineInTraceback */

Import */

Import */