Skip to content

v0.2.0

Latest

Choose a tag to compare

@okunator okunator released this 20 Oct 16:10

Upgrade notes

  • Recheck training and evaluation baselines. Corrected losses change values and
    gradients; reconstruction and metric fixes can change masks and scores.
    Network checkpoint shapes and names remain compatible unless noted below.
  • InstanSeg training now uses the shared classifier:
    Loss("instanseg", pixel_classifier=network.probability_net). Optimize the
    network, including that classifier. Replace w_coord with w_inst and retrain
    models trained only with the old coordinate-MSE loss.
    See the loss API.
  • MSSSIM requires matching floating BCHW maps on the same device and dtype;
    its default window needs at least 161 pixels per axis. Set data_range
    (default 1). For standalone SSIM, floating BHW targets are shared maps;
    integer BHW targets are class labels. Set its range explicitly for comparisons.
  • JointLoss requires one weight per loss; MultiTaskLoss requires weight keys
    matching configured heads. Invalid configurations now raise errors.
  • Prefer CellVit's hgsc_v1_samvit_base_patch16 checkpoint name; the old
    hgsc_v1_efficientnet_b5 remains an alias. Enable the
    installation extras for training
    datasets, ONNX export or OpenSlide.

Removed

  • Remove CPPNet and its checkpoints/export because borrowed sampling code lacks
    an established redistribution license. Other model families remain available.
  • Remove Linformer attention and DCAN/DRFNS reconstruction, including DCAN head
    names. Use exact attention and a supported model/reconstruction pair.
    Changing from Linformer requires retraining, not checkpoint conversion.
  • Remove datasets.Pannuke, utils.Downloader and
    FileHandler.extract_zips_in_folder. Use the README's
    PanNuke download example, Hugging Face tools and
    Python's zipfile.ZipFile. Training loaders remain in torch_datasets.

Features

  • Add training, onnx and openslide installation extras.
  • Add ONNX export for CellPose, StarDist, HoVer-Net, CellVit and InstanSeg;
    document Omnipose export through export_cellpose_onnx. Exports support fixed
    spatial dimensions and optional variable batches; instance reconstruction stays
    in Python. CellPose/StarDist export was contributed by Ajinkya Kulkarni.
  • Add calibrated SlideReader.pixel_size in (height, width) micrometers and
    optional cuCIM CUDA reads with WsiSegmenter(read_device="cuda"). Decoded tiles
    stay on GPU through normalization and model input; reconstruction and vector
    export remain on CPU. CPU reads remain the default.
  • Add vector-only nucleus filtering after WSI stitching. filter_level 1–3
    selects conservative, middle or strong area/thickness limits with physical
    pixel calibration; 0 or None disables filtering by default.

Fixes

  • Correct Dice, IoU and Tversky spatial overlap, masking, weighting and spectral
    regularization. Fix CE/Focal and BCE/SCE logit stability, BCE channel weighting
    and binary smoothing, and MSE/MAE mixed-precision arithmetic.

  • Correct MS-SSIM scale aggregation: identical maps now have zero loss. Stabilize
    standalone SSIM precision, gradients and target handling; add range control.

  • Align InstanSeg's shared-classifier membership, seed and instance losses with
    reconstruction. Fix loss routing, spatial/uniform smoothing and weighted
    multi-head sums; accept target-only edge-weight maps in MultiTaskLoss.

  • Prevent CellVit float16 upsampling overflow, fixed-unpool bfloat16 dtype errors
    and NaNs from zero-norm style vectors. Convert bfloat16 maps to float32 before
    NumPy post-processing. Load StarDist ray counts from checkpoint weights.

  • Reconstruct HoVer-Net/CellVit nuclei from binary foreground rather than class
    numbers, retaining cell types for voting.

  • Reconcile overlapping WSI polygons in vector space, preferring complete nuclei
    from one tile while preserving unmatched detections and distinct overlapping
    nuclei. Avoid artificial intersection objects, disconnected seam fragments and
    coordinate-sensitive selection. Preserve IDs and attributes; reject existing
    merge outputs and repair invalid contours after smoothing.

  • Correct slide height/width scale ordering and level-zero polygon coordinates,
    cuCIM pyramid scaling/padding, tile inference arguments and optional output
    handling. Preserve foreground class 0 during vector export.

  • Fix in-memory sliding-window padding, overlap normalization, boundary weighting
    and optional maps; reject strides that leave gaps. Apply requested predictor
    boundary weights to reconstruction and foreground maps.
    (#69)

  • Correct one-to-one instance matching below 0.5 IoU, swapped object FP/FN counts,
    multiclass confusion counts and specificity. Support sparse IDs and masks
    without background; retain absent-class handling and historical AJI matching
    and average_precision semantics.
    (#73)

  • Keep auxiliary training targets aligned after augmentation, apply all sampled
    StrongAugment operations, and correct HDF5 lengths and multi-mask target routing.

  • Use built-in min-max normalization in the quick start with float32 output
    (#78). Honor
    percentile clipping bounds and avoid overflow for extreme finite min-max inputs.
    Correct the Normalization export.

  • Support Python 3.10–3.14, excluding 3.14.1 per torchvision. Refresh dependencies
    with minimum-only runtime requirements, including NumPy 2.2.6+ and PyTorch
    2.13+. Use headless OpenCV and leave transitive constraints to parent packages.
    (#68)

  • Allow inference imports without optional training or plotting packages.
    Bundle third-party licenses and exclude tests/fixtures from distributions.