Thanks for your release. I use this repo to replicate the FSM result, but it can not replicate it at all, RMSE is more than 170 without gt scale. Have you tested before release?
this is my config setting:
wrapper:
recipe: wrapper|default
max_epochs: 30
validate_first: False
arch:
model:
file: depth/FSMModel
networks:
depth:
recipe: networks/focal_depth_res_net|fsm_ddad
depth_range: [0.1,200.0]
pose:
recipe: networks/conv_pose_net|default
losses:
reprojection:
recipe: losses/reprojection|default
smoothness:
recipe: losses/smoothness|default
evaluation:
depth:
recipe: evaluation/depth|ddad_resize
optimizers:
depth:
recipe: optimizers|adam_20_05
pose:
recipe: optimizers|adam_20_05
datasets:
train:
recipe: datasets/ddad|train_selfsup_6cams
dataloader:
batch_size: 1
validation:
recipe: datasets/ddad|validation_6cams
Thanks for your release. I use this repo to replicate the FSM result, but it can not replicate it at all, RMSE is more than 170 without gt scale. Have you tested before release? this is my config setting: wrapper: recipe: wrapper|default max_epochs: 30 validate_first: False arch: model: file: depth/FSMModel networks: depth: recipe: networks/focal_depth_res_net|fsm_ddad depth_range: [0.1,200.0] pose: recipe: networks/conv_pose_net|default losses: reprojection: recipe: losses/reprojection|default smoothness: recipe: losses/smoothness|default evaluation: depth: recipe: evaluation/depth|ddad_resize optimizers: depth: recipe: optimizers|adam_20_05 pose: recipe: optimizers|adam_20_05 datasets: train: recipe: datasets/ddad|train_selfsup_6cams dataloader: batch_size: 1 validation: recipe: datasets/ddad|validation_6cams