tensor image: (3, H, W)

Computation of error metrics between predicted and ground truth depths Args: gt (N): ground truth depth pred (N): predicted depth

depth: (H, W)

loss for a given set of pixels: z_A: predicted absolute depth for pixels A z_B: predicted absolute depth for pixels B

Convert euler angles to rotation matrix. Reference: https://github.com/pulkitag/pycaffe-utils/blob/master/rot_utils.py#L174 Args: ang

evaluate depth result Args: gt_depths: list of gt depth files pred_depths: list of predicted depth files e

Computes the smoothness loss for a disparity image The color image is used for edge-aware smoothness

Inverse warp a source image to the target image plane. Args: img: the source image (where to sample pixels) -- [B, 3, H, W] d

Convert 6DoF parameters to transformation matrix. Args:s vec: 6DoF parameters in the order of tx, ty, tz, rx, ry, rz -- [B, 6] Re

Convert quaternion coefficients to rotation matrix. Args: quat: first three coeff of quaternion of rotation. fourht is then computed to ha

Constructs a ResNet model. Args: num_layers (int): Number of resnet layers. Must be 18 or 50 pretrained (bool): If True, returns a

Upsample input tensor by a factor of 2

inputs and targets: surface normal image images: rgb images