0.7.2

What's Changed

• Update location to datasets submodule by @anwai98 in #245
• Update ViM-UNet documentation by @anwai98 in #246
• Change instance norm settings in U-Net to avoid boundary artifacts by @constantinpape in #247
• Add SIIM ACR Pneumothorax dataset by @anwai98 in #256
• Add CAMUS dataset by @anwai98 in #257
• Add DRIVE dataset by @anwai98 in #260
• Add Papila dataset by @anwai98 in #262
• Update torch_em datasets by @constantinpape in #252
• Add BUSI dataset by @anwai98 in #268
• Update dataset doc path by @anwai98 in #265
• Add stuff for doc by @constantinpape in #269
• Add Plethora dataset by @anwai98 in #263
• Update default optimizer to AdamW by @anwai98 in #284
• Add SA-Med2D-20M dataset by @anwai98 in #282
• Add Medical Segmentation Decathlon dataset by @anwai98 in #224
• Add OSIC Pulmonary Fibrosis dataset by @anwai98 in #281
• Minor update SA-Med2d 20M dataset by @anwai98 in #290
• Add Montgomery dataset by @anwai98 in #249
• Add IDRiD dataset by @anwai98 in #266
• Add UWaterloo Skin dataset by @anwai98 in #261
• Minor updates to CAMUS dataset by @anwai98 in #292
• Add OrgaSegment dataset by @anwai98 in #231
• Add FeTa dataset by @anwai98 in #293
• Add OIMHS dataset by @anwai98 in #291
• Add JNU-IFM dataset by @anwai98 in #270
• Add MicroUSP dataset by @anwai98 in #279
• Add Seg-A dataset by @anwai98 in #276
• Multi gpu training by @constantinpape in #283
• Create splits for OIMHS data by @anwai98 in #300
• Restore old norm layer defaults by @constantinpape in #303
• Update resize functionality for SIIM ACR by @anwai98 in #308
• Minor update to preprocessing logic for Papila by @anwai98 in #307
• Add resizing to OSIC PulmoFib dataset by @anwai98 in #306
• Add CholecSeg8k dataset by @anwai98 in #305
• Add Duke Liver dataset by @anwai98 in #304
• Add PICCOLO dataset by @anwai98 in #302
• Add CBIS DDSM dataset by @anwai98 in #301
• Add m2caiseg dataset by @anwai98 in #280
• Add DCA1 dataset by @anwai98 in #309
• Add ISIC dataset by @anwai98 in #250
• Minor fix to UNETR mini-batch norm by @anwai98 in #310
• Add AMOS dataset by @anwai98 in #255
• Add Toothfairy dataset by @anwai98 in #313
• Minor fix to adapt label conversion scripts for papila by @anwai98 in #312
• Minor updates to drive dataset by @anwai98 in #314
• Minor update to SegA data structure by @anwai98 in #316
• Add SPIDER dataset by @anwai98 in #315
• Add HaN Seg dataset by @anwai98 in #275
• Add ACDC dataset by @anwai98 in #251
• Add Covid19 Lung and Infection Segmentation Dataset by @anwai98 in #253
• Add organoidnet dataset by @constantinpape in #318
• Cellseg 3d by @constantinpape in #319
• added label transform class encorporating min_size of labels by @lufre1 in #322
• Bump version by @constantinpape in #321

New Contributors

• @lufre1 made their first contribution in #322

Full Changelog: 0.7.1...0.7.2

0.7.1

What's Changed

Support for updated bioimage.io libraries:

• Update links to the vimunet benchmarking by @anwai98 in #239
• Minor touch-up updates to torch-em docs by @anwai98 in #241
• Add random colors generator for instance plotting by @anwai98 in #244
• Add documentation on datasets and dataloaders by @anwai98 in #234
• Bump bioimageio dependency by @constantinpape in #243

Full Changelog: 0.7.0...0.7.1

0.7.0

What's Changed

• Fix missing attribute for dicelosswithlogits by @anwai98 in #220
• Minor fix to ctc exception message by @anwai98 in #221
• Minor update to sort encoder outputs by @anwai98 in #219
• Add DynamicNuclearNet data - from DeepCell by @anwai98 in #228
• Add top-level ViM-UNet description file by @anwai98 in #235
• Add ViM-UNet by @anwai98 in #236
• Set DynamicNuclearNet download default to False by @anwai98 in #230
• Add ASEM dataset by @anwai98 in #226
• Enable passing external pbar to trainer by @constantinpape in #232
• Bump version by @constantinpape in #238

Full Changelog: 0.6.2...0.7.0

0.6.2

What's Changed

• Implement grid-search for distance based segmentation by @constantinpape in #193
• Update the sampling mechanism in the image collection dataset by @constantinpape in #195
• Fix AxonDeepSeg by @anwai98 in #196
• Bump Version by @anwai98 in #197
• Update NucMM dataset by @anwai98 in #199
• Fix mitoem rat volume download by @anwai98 in #202
• Minor update to UNETR by @anwai98 in #203
• Minor fix to covid-if data description by @anwai98 in #206
• Update grid-search by @constantinpape in #209
• Update ISBI dataset download by @anwai98 in #210
• Minor fix to NeurIPS CellSeg dataset by @anwai98 in #212
• Update UNETR upsampling by @anwai98 in #211
• Add BTCV dataloader by @anwai98 in #171
• Add CTC dataset by @constantinpape in #214
• Add AutoPET dataset by @anwai98 in #213
• Refactor CTC dataset by @anwai98 in #215
• Update UNETR - making the training modular by @anwai98 in #217
• Add test set for CTC dataset by @anwai98 in #216
• Add current_metric while saving checkpoints by @anwai98 in #223

Full Changelog: 0.6.1...0.6.2

0.6.1

What's Changed

• Implement different channel reductions for dice by @constantinpape in #194

Full Changelog: 0.6.0...0.6.1

Transformer based segmentation functionality

What's Changed

• Implement UNETR by @constantinpape in #146
• Fixes and Updates for UNETR by @anwai98 in #148
• Minor Fix to UNETR Decoder by @anwai98 in #149
• Add imagecodecs to the dependencies to fix CI by @constantinpape in #150
• Add PanNuke Dataloader by @anwai98 in #153
• Refactor UNETR Training by @anwai98 in #151
• Add Monai's UNETR by @anwai98 in #155
• Add BCSS Dataset by @anwai98 in #156
• Add PanNuke Check Script by @anwai98 in #159
• Updates to MoNuSeg by @anwai98 in #157
• Add Watershed Segmentation Functionality by @anwai98 in #160
• Add MoNuSAC Dataset by @anwai98 in #158
• Raise download prompt - NeurIPS CellSeg by @anwai98 in #161
• Raise warning for Lizard data download by @anwai98 in #163
• Add type annotations - PanNuke by @anwai98 in #165
• Add splits for BCSS dataset by @anwai98 in #164
• UNETR Training with Affinities by @anwai98 in #166
• Save train time in checkpoint by @constantinpape in #169
• Add Raw Transform - nnUNet by @anwai98 in #170
• Bring main up-to-date by @constantinpape in #173
• Add ViT-MAE and Refactor ViT(s) by @anwai98 in #172
• Distance label trafos by @constantinpape in #177
• Fix UNETR Model for LIVECell by @anwai98 in #174
• Add LIVECell Distance Map Experiments by @anwai98 in #175
• Add distance based loss by @constantinpape in #178
• Fix issues in ViT initialization and update UNETR state loading by @constantinpape in #180
• Misc updates to transforms, loss and UNETR by @constantinpape in #181
• Fix UNETR Training - Update Rescaling Inputs by @anwai98 in #179
• Sync dev by @constantinpape in #184
• WIP - Instance Segmentation Experiments by @anwai98 in #187
• Dev by @constantinpape in #188
• Update DeepBacs Loader by @anwai98 in #189
• Implement mitolab benchmark datasets by @constantinpape in #190
• Update UNETR to enable resize to longest side by @constantinpape in #192

Full Changelog: 0.5.2...0.6.0

0.5.2

To fix conda build.

0.5.1

What's Changed

• Update Teacher Initialization for MT Setup by @anwai98 in #132
• TissueNet Dataloader by @anwai98 in #134
• Add deepbacs dataloader by @constantinpape in #135
• Implement get_dataset for more datasets by @constantinpape in #137
• Update check_loader to save plots by @anwai98 in #136
• Dataset updates by @constantinpape in #138
• Implement NucMM Dataset by @constantinpape in #139
• Implement sponge em dataset by @constantinpape in #142
• Add doc strings to datasets to provide the correct dataset references by @constantinpape in #143
• Refactor experiments by @constantinpape in #144
• Update trainer.fit by @constantinpape in #145

Full Changelog: 0.5.0...0.5.1

Classification and domain adaptation functionality

What's Changed

• Implement MeanTeacher Trainer by @constantinpape in #112
• Add support for torch2 by @constantinpape in #113
• Update Prob-DA Source UNet Parameters by @anwai98 in #115
• New datasets and other changes by @constantinpape in #114
• Load utils by @constantinpape in #117
• Update Mean-Teacher and FixMatch Self-Training Scheme(s) by @anwai98 in #116
• Add all necessary tests for mean teacher trainer deserialization by @constantinpape in #118
• Add 2D Probabilistic UNet by @anwai98 in #119
• Add Argument - Target Domain by @anwai98 in #121
• Update Evaluation with Distribution Alignment by @anwai98 in #122
• Add 2D Probabilistic UNet Training Setup by @anwai98 in #123
• Add Probabilistic Self-Training (MT) Approaches by @anwai98 in #125
• Add FixMatch-based Probabilistic Domain Adaptation Setups by @anwai98 in #126
• Make livecell label extraction more robust against duplicate per obje… by @constantinpape in #127
• Add CLI for UNet training and prediction by @constantinpape in #128
• Implement tif support in data loaders by @constantinpape in #130
• Update Prediction Functions for PUNet Domain Adaptation Setups by @anwai98 in #129
• Implement classification functionality by @constantinpape in #120

Full Changelog: 0.4.1...0.5.0

0.4.1

What's Changed

• fix best_metric < worst_metric by @FynnBe in #50
• refactor _train_epoch and _validate by @FynnBe in #51
• refactor DefaultTrainer.from_checkpoint by @FynnBe in #53
• Deserialize ROI Wrapper WIP by @constantinpape in #55
• fix error connected to retain_grad() by @JonasHell in #57
• Add Serializer to DefaultTrainer by @FynnBe in #54
• fix typo in augmentations by @JonasHell in #59
• Add sampler that checks if there are a minimum number of instances in the volume by @JonasHell in #60
• ensure correct behaviour of filters when set manually by @JonasHell in #61
• Add functionality to mask to-bg-boundaries by @JonasHell in #62
• Implement advanced rf training for s2d by @constantinpape in #63
• Update s2d features, extend mito-enhancer training by @constantinpape in #64
• Enable overwriting the min-shape in modelzoo export by @constantinpape in #65
• Extend shallow2deep experiments by @constantinpape in #67
• S2D experiments for training LM enhancers by @constantinpape in #71
• add noise augmentations for mean teacher approach by @JonasHell in #68
• fix typo/bug in validation logging by @JonasHell in #73
• Add axon myelin training by @constantinpape in #74
• add worst_tiles sampling by @JonasHell in #77
• Extend s2d functionality by @constantinpape in #75
• Elastic 3d by @Buglakova in #79
• adapt default mean teacher augs to work with modelzoo export by @JonasHell in #82
• adapt _serialize_feature_config() to work with strings by @JonasHell in #83
• Add tests for augmentations by @constantinpape in #80
• Add distance transform to label transforms by @constantinpape in #84
• Add save_root argument to trainer to enable saving files at different… by @constantinpape in #85
• Use ApplyMask in other masking loss wrappers by @constantinpape in #86
• Make bioimageio import optional by @constantinpape in #88
• fix bioimageio import in shallow2deep by @JonasHell in #89
• Use tifffile to read compressed tif files by @Czaki in #93
• When loading data for train allow to read data that has channel on first position by @Czaki in #94
• Implement sampler for the image collection dataset by @constantinpape in #96
• provide reasonable default values for augmentations by @JonasHell in #97
• Add resnet 3d implementation (based on torchvision) by @constantinpape in #99
• Update livecell.py by @constantinpape in #100
• Augmentations by @JonasHell in #98
• Update spoco losses by @constantinpape in #87
• implement gradient masking with cropping - new PR by @JonasHell in #101
• Add Separate Label Path to the LiveCELL Loader by @anwai98 in #104
• Update Label Type Argument for LiveCELL by @anwai98 in #105
• Raise value error if dice is called with non-matching shapes by @constantinpape in #106
• add args to adjust tqdm output by @JonasHell in #108
• add dense accumulate as additional baseline by @JonasHell in #107
• Add neurips cell seg dataset; update raw image collection dataset; up… by @constantinpape in #111
• Update Checksum for UroCell Dataloader by @anwai98 in #109
• Misc changes WIP by @constantinpape in #110

New Contributors

• @Buglakova made their first contribution in #79
• @Czaki made their first contribution in #93
• @anwai98 made their first contribution in #104

Full Changelog: 0.4.0...0.4.1