Skip to content

Latest commit

 

History

History
159 lines (78 loc) · 4.5 KB

CHANGELOG.md

File metadata and controls

159 lines (78 loc) · 4.5 KB
Raw

Release notes

0.2.1

-- Bug fixes -- Adding Code-based Tutorial (https://matjesg.github.io/deepflash2/tutorial_monuseg.html)

0.2.0

New Features

  • new features (#42), thanks to @matjesg

    • Multiclass GT Estimation, closes #34
    • Torchscript ensemble class for inference / tta adjusted
    • ONNX export possible

  • Major changes

    • Different classes for training (EnsembleLearner) and Inference (EnsemblePredictor)
    • Normalization based on uin8 images (0...255)

  • Instance segmentation metrics (#32), thanks to @matjesg

    • Instance segmentation metrics for GT estimation and prediction

0.1.8

  • Dependency fix (opencv-python-headless>=4.1.1,<4.5.5)
  • Minor improvements and fixes

0.1.7

  • Fixes bug when predicting on new data.

0.1.6

  • Removing git dependency for cellpose installation.

0.1.5

  • Bug fixes and minor improvements.

0.1.4

New Features

  • Instance segmentation metrics (#32), thanks to @matjesg
      • Instance segmentation metrics for GT estimation and prediction
  • Tutorials

0.1.3

New Features

  • Adding tifffile imread (#13), thanks to @matjesg

    • Closes #12

  • Multichannel tiffs are imported as single channel greyscale when imported via ui (#12)

    • Problem: .tiff images are with shape z,c,y,x = 1,4,1024,2048 are selected in the user deepflash google colab user interface. Image preview only shows first channel. After inspecting the image import, I saw, that the tiffile has dimensions 1024x2048 after import.

Proposed solution: Images may be imported with tifffile.imread instead of imageio.imread. Here, the channel dimensions are preserved.

0.1.2

New Features

  • Real-Time loss weight computation and large file support (#11), thanks to @matjesg
      • Real-Time loss weight computation via fast convolutional distance transform on GPU
  • temp storage using zarr instead of RAM
  • zarr dependency

Bugs Squashed

  • Prediction not possible on Windows machines due to cuda error (#10)
    • During prediction on a windows machine a a OS Error occurs due to:

"RuntimeError: cuda runtime error (801) : operation not supported at ..\torch/csrc/generic/StorageSharing.cpp:247"

Problem: Storage sharing currently not supported on windows.

Proposed solution: Ensemble learner takes "num_workers" argument and passes it to subsequent functions. If num_workers == 0, prediction works for me.

0.1.1

  • Bug fixes and minor improvements.

0.1.0

New Features

  • Adding GUI and new project structure (#8), thanks to @matjesg

0.0.14

New Features

  • Adding test time augmentation (#7), thanks to @matjesg

Bugs Squashed

  • Type Error when starting training (#5)
    • When I try to start the training process in google colab, this error occurs:

TypeError: no implementation found for 'torch.nn.functional.cross_entropy' on types that implement torch_function: [<class 'fastai.torch_core.TensorImage'>, <class 'fastai.torch_core.TensorMask'>]

as well as

FileNotFoundError: [Errno 2] No such file or directory: 'models/model.pth'

in the end.

Hope you know whats the problem.

0.0.13

Bugs Squashed

  • Type Error when starting training (#5)
    • When I try to start the training process in google colab, this error occurs:

TypeError: no implementation found for 'torch.nn.functional.cross_entropy' on types that implement torch_function: [<class 'fastai.torch_core.TensorImage'>, <class 'fastai.torch_core.TensorMask'>]

as well as

FileNotFoundError: [Errno 2] No such file or directory: 'models/model.pth'

in the end.

Hope you know whats the problem.

0.0.12

Bugs Squashed

  • Checking for SimpleITK 1.2.4 and install if not available (#4), thanks to @matjesg
    • Closes #3