sum over the first and last dimention

Returns the model's average precision for each class Return: ap (FloatTensor): 1xK tensor, with avg precision for each class k

1x1 convolution

3x3 convolution with padding

3x3 convolution with padding

Execute an replication callback `__data_parallel_replicate__` on each module created by original replication. The callback will be invoked wi

+-----+-----+-----+-----+ | 1 | 2 | 3 | 4 | +-----+-----+-----+-----+ +-----+-----+ | 1 | 2 | +-----+-----+

Constructs a ResNet-101 model. Args: pretrained (bool): If True, returns a model pre-trained on ImageNet :param pretrained:

Constructs a ResNet-152 model. Args: pretrained (bool): If True, returns a model pre-trained on ImageNet

Constructs a ResNet-18 model. Args: pretrained (bool): If True, returns a model pre-trained on ImageNet

Constructs a ResNet-34 model. Args: pretrained (bool): If True, returns a model pre-trained on ImageNet

Constructs a ResNet-50 model. Args: pretrained (bool): If True, returns a model pre-trained on ImageNet :param pretrained:

Args: master_callback: a callback to be invoked after having collected messages from slave devices.

Compute the mean and standard-deviation with sum and square-sum. This method also maintains the moving average on the master device.

Load the image and label in numpy.ndarray

Load the image and label in numpy.ndarray

Create a file path/image id list.

Create a file path/image id list.

image: (3, H, W) am: (H/f, W/f) range(0,1)

image: (3, H, W) am: (H/f, W/f) range(0,1)

CRF post-processing on pre-computed logits

x: numpy array

x: numpy array

Evaluation on test set