porwalnaman01
commented
3 years ago Hello there, did you solve this issue? I am also using ImageDataset but when I try to run the script, it reads the validation files but do not read the training files. My dataset is a folder containing .jpg images. Thanks in advance!
Config file that I am using :
model:
name: 'SelfSupModel'
optimizer:
name: 'Adam'
depth:
lr: 0.0002
pose:
lr: 0.0002
scheduler:
name: 'StepLR'
step_size: 30
gamma: 0.5
depth_net:
name: 'DepthResNet'
version: '50pt'
pose_net:
name: 'PoseNet'
version: ''
params:
crop: 'garg'
min_depth: 0.0
datasets:
augmentation:
image_shape: (192, 640)
train:
batch_size: 4
dataset: ['Image']
path: ['/disk1/dan/datasets/vgg-faces/train']
split: ['train_split.txt']
validation:
dataset: ['Image']
path: ['/disk1/dan/datasets/vgg-faces/val']
split: ['val_split.txt']
checkpoint:
filepath: '/disk1/dan/Naman/packnet-sfm-0.1.2/experiments1'
monitor: 'abs_rel_pp_gt'
monitor_index: 0
mode: 'min'
I am trying to finetune model for my own dataset and using ImageDataset as dataset class. Could you please explain what is split file in .yaml file? For example, I have downloaded KITI tiny and in there for image_02 images you use corresponding in image_03. They look like real stereo examples, since images are not equal . So the question: What is correct way to make a split file? Is it correct to make the same file as in KITI_tiny with monocular camera: image_1 image_2 image_2 image_3 ........................... imagen image{n+1}