nuclear_speckles_analysis

This repository trains a UNet-style image-to-image translation model on cropped nuclei from multiple nuclear speckle datasets (including the initial dataset and U2OS).

• Input: cropped DAPI nucleus image
• Target: cropped Gold nucleus image
• Task: predict Gold crops from DAPI crops

Dataset-specific channel mappings:

• initial: DAPI=CH0, Gold=CH2
• u2os: DAPI=CH01, Gold=CH03

Dataset-specific cache roots:

• initial: /mnt/big_drive/nuclear_speckle_data/initial_dataset/model_cache
• u2os: /mnt/big_drive/nuclear_speckle_data/u20s_dataset_jan_15_2026/model_cache

Within each cache root, training uses:

• dapi_to_gold_crop_cache
• paired_tensor_cache

Normalization

Before samples are passed through the model, both input and target crops are normalized by their channel max intensity (dtype max, for example 255 for uint8 and 65535 for uint16) in datasets/dataset_00/utils/ImagePreProcessor.py.

Training

Main entrypoint:

mlflow run . -e train_model

Fast smoke run:

uv run train.py --epochs 1 --n-trials 1 --max-train-batches 2 --max-eval-batches 2 --enable-image-savers 0

# Select dataset (defaults to u2os)
uv run train.py --dataset initial --epochs 1 --n-trials 1 --max-train-batches 2 --max-eval-batches 2 --enable-image-savers 0

Loss and Metrics

• Optimization loss: L1
• Logged metrics: L1, L2, PSNR, SSIM

Metrics are implemented in metrics/ and logged through callbacks/Callbacks.py.

Outputs

• Crop-level prediction artifacts can be logged each epoch via callbacks/utils/SaveEpochCrops.py.
• Whole-FOV reconstruction and patch stitching are intentionally removed.