Small adaptation layers.

Early stops the training if validation loss doesn't improve after a given patience.

Small adaptation layers.

Small adaptation layers.

DFNet implementation

A slight accelerated version of DFNet, we experimentally found this version's performance is similar to original DFNet but inferences faster

DFNet with EB3 backbone

DFNet with EB0 backbone, feature levels can be customized

EfficientNet-B3 backbone, model ref: https://github.com/lukemelas/EfficientNet-PyTorch/blob/master/efficientnet_pytorch/model.py

EfficientNet-B3 backbone, model ref: https://github.com/lukemelas/EfficientNet-PyTorch/blob/master/efficientnet_pytorch/model.py

simple L_2-Loss

Implements the MapNet model (green block in Fig. 2 of paper)

Equation 13 in the NeRF-W paper. Name abbreviations: c_l: coarse color loss f_l: fine color loss (1st term in equation 13)

add skip connections

vgg encoder

vgg encoder with bilinear upsampling

vgg encoder with bilinear upsampling

vgg encoder with bilinear upsampling

vgg encoder with bilinear upsampling