Open valosekj opened 5 months ago
I tried three experiments, all for 500 epochs, modifying:
batch_size=8
, 5 holes
Reconstructed images across epochs: link
👉 visually looks the best, training now for more epochs
batch_size=8
, 10 holes
Reconstructed images across epochs: link
batch_size=16
, 10 holes
Reconstructed images across epochs: link
Two experiments with RandCoarseDropoutd augmentation transform: fill_value=None
vs fill_value=0
. Discussion here.
Both used batch_size=8, 5 holes, and 1500 epochs.
RandCoarseDropoutd fill_value=None
= filling the dropped regions with randomly selected value based on the min and max value of input image
Reconstructed images across epochs: link
RandCoarseDropoutd fill_value=0
= filling the dropped regions with value 0
Reconstructed images across epochs: link
dcm-zurich-lesions
Since the pre-training seems to work (at least the code does not raise errors and the model is learning to reconstruct the images), I performed two fine-tuning experiments: using a pre-trained model vs training from scratch.
Fine-tuning using the SSL pre-trained model with RandCoarseDropoutd fill_value=0
(see comment above).
Although the learning crashed, the model was learning something! See the validation figure below.
Fine-tuning from the scratch (weights were initialized randomly).
This issue summarizes the follow-up experiment to https://github.com/ivadomed/model-seg-dcm/issues/7.
Pre-training
In https://github.com/ivadomed/model-seg-dcm/issues/7., the entire input image of size (64, 256, 256) was used to create two augmeted views. Here, two random samples of size (64, 64, 64) are taken from the input image to create the augmented view using these transforms. Moreover, these samples are taken around a label (spinal cord mask) using RandCropByPosNegLabeld: