PoseNet

class packnet_sfm.networks.pose.PoseNet.PoseNet(nb_ref_imgs=2, rotation_mode='euler', **kwargs)

Bases: torch.nn.modules.module.Module

Pose network

forward(image, context)

Defines the computation performed at every call.

Should be overridden by all subclasses.

Note

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the registered hooks while the latter silently ignores them.

init_weights()

packnet_sfm.networks.pose.PoseNet.conv_gn(in_planes, out_planes, kernel_size=3)

Convolutional block with GroupNorm

Parameters

• in_planes (int) – Number of input channels

• out_planes (int) – Number of output channels

• kernel_size (int) – Convolutional kernel size

Returns

layers – Sequence of Conv2D + GroupNorm + ReLU

Return type

nn.Sequential