Configurable identity-mapping residual block Parameters ---------- in_channels : int Number of input channels.

Wider ResNet with pre-activation (identity mapping) blocks Parameters ---------- structure : list of int Number o

This is the initialization for class uniform sampling :param size: crop size (int) :param crop_nopad: Padding or no padding (

Init :param params: parameters to optimize :param lr: learning rate :param betas: beta :param eps: numerical

Fill a 2D tensor with a constant scale value

Single Layer BN and Relui

The first step of uniform sampling is to decide sampling centers. The idea is to divide each image into tiles and within each tile, we co

Generate an epoch of crops using uniform sampling. Needs to be called every epoch. Will not apply uniform sampling if not train or class

Evaluates module(input) in parallel across the GPUs given in device_ids. This is the functional version of the DataParallel module. Args:

Compute F score for a segmentation mask Arguments: seg_mask (ndarray): segmentation mask prediction gt_mask (ndarray): segme