MS-SSIM Joint Training Loss#34
Closed
MattsonCam wants to merge 55 commits into
Closed
Conversation
Experiment ID:2 Run ID:5e9a052521da4320871db9625e09cf78 Run Name:agreeable-turtle-117 - ConvNeXtUNet Generator - Single 2D crop input and single 2D crop target - Cache-backed filtered nucleus crops generated from the configured data directory - L1 optimization objective with L2, PSNR, SSIM, and Pearson correlation metric logging - Normalization of input and target images using the max possible pixel intensity - Training on entire initial dataset - Sigmoid output activation function when computing metrics and loss and saving images
Generator Model Training Orchestration
reduced crop size from training orchestration
Resampled Nuclear Speckle U2OS Dataset
Experiment ID:3 Run ID:763e5f09e64b455ba6981cfdf5fa60d7 Run Name:bittersweet-tern-183 Optimization of a DAPI-to-Gold image-to-image translation model with: - ConvNeXtUNet Generator - Single 2D crop input and single 2D crop target - Cache-backed filtered nucleus crops generated from the configured data directory - Train-split z-score normalization for inputs and targets - L1 optimization objective in z-score space with denormalized L2, PSNR, SSIM, and Pearson correlation metric logging - Resampling of dataset w/o anti-aliasing -> dtype clipping (to stay within the correct range) -> z-score normalization - Means and stds from the training set were stored for z-scoring of the validation set - No final activation function after the model output (such as sigmoid)
Experiment ID:3 Run ID:4feeef6283384834b907e9b1af611524 Run Name:aged-dove-229 Optimization of a DAPI-to-Gold image-to-image translation model with: - ConvNeXtUNet Generator - Single 2D crop input and single 2D crop target - Cache-backed filtered nucleus crops generated from the configured data directory - Train-split z-score normalization for inputs and targets - L1 optimization objective in z-score space with denormalized L2, PSNR, SSIM, and Pearson correlation metric logging - Resampling of dataset w/o anti-aliasing -> dtype clipping (to stay within the correct range) -> z-score normalization - Means and stds from the training set were stored for z-scoring of the validation set - No final activation function after the model output (such as sigmoid)
Refactor normalization, evaluation, and preview generation for z-scored training
Run ID:5cbf036ef8284509850ab17794c3271a Run Name:amusing-fly-350 Optimization of a DAPI-to-Gold image-to-image translation model with: **Note**: This run may have included uncentered cached crops once transformed - ConvNeXtUNet Generator - Single 2D crop input and single 2D crop target - Cache-backed filtered nucleus crops generated from the configured data directory - Train-split z-score normalization for inputs and targets - L1 plus Optuna-weighted SSIM optimization objective in z-score space with denormalized L2, PSNR, SSIM, and Pearson correlation metric logging - Resampling of dataset w/o anti-aliasing -> dtype clipping (to stay within the correct range) -> z-score normalization - Means and stds from the training set were stored for z-scoring of the validation set - No final activation function after the model output (such as sigmoid)
Experiment ID:3 Run ID:8357552379084e8bbada22b6a7b0561e Run Name:treasured-flea-870 - ConvNeXtUNet Generator - Single 2D crop input and single 2D crop target - Cache-backed filtered nucleus crops generated from the configured data directory - Train-split z-score normalization for inputs and targets - L1 plus Optuna-weighted SSIM optimization objective in z-score space with denormalized L2, PSNR, SSIM, and Pearson correlation metric logging - Resampling of dataset w/o anti-aliasing -> dtype clipping (to stay within the correct range) -> z-score normalization - Means and stds from the training set were stored for z-scoring of the validation set - No final activation function after the model output (such as sigmoid)
Robust Percentile Normalization
Experiment ID:3 Run ID:b04b5d86bcea40cc93ca0e5f0cf17d6e Run Name:stylish-turtle-260 - ConvNeXtUNet Generator - Single 2D crop input and single 2D crop target - Cache-backed filtered nucleus crops generated from the configured data directory - L1 plus Optuna-weighted SSIM optimization objective in z-score space with denormalized L2, PSNR, SSIM, and Pearson correlation metric logging - Resampling of dataset w/o anti-aliasing -> dtype clipping (to stay within the correct range) -> z-score normalization - Means and stds from the training set were stored for z-scoring of the validation set - No final activation function after the model output (such as sigmoid) - Train-split 1st/99th percentile normalization for inputs and targets - L1 plus Optuna-weighted SSIM optimization objective in normalized space with denormalized L2, PSNR, SSIM, and Pearson correlation metric logging
|
Check out this pull request on See visual diffs & provide feedback on Jupyter Notebooks. Powered by ReviewNB |
Member
Author
|
No longer necessary because of #35 |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
This pr updates the training objective to use a 4-scale MS-SSIM term alongside L1 loss for 128x128 normalized images.