v4.2.1.1 new Cellpose4 models!

Check out the new Cellpose4 models, described here.

What's Changed

• Remove .views & .color states, fix gradZ 3D GUI display by @mrariden in #1432
• Resample fix by @mrariden in #1438
• Imagedraw fixes & update logging by @mrariden in #1431
• allow errors to propagate past Qt, add debug logging by @mrariden in #1448
• speed up w/ flattened inds instead of tuple (claude suggestion) and #… by @carsen-stringer in #1447
• Fix model removal from GUI by @mrariden in #1440
• Fix npy channel by @mrariden in #1441
• fix merge with background error and add debug logging by @mrariden in #1449
• Escape drawing, move reset view to space, disable right click zoom by @mrariden in #1451
• adding dino models and gpu-accelerated augmentations by @carsen-stringer in #1454
• Fix sat by @carsen-stringer in #1461
• adding doc changes for dino models by @carsen-stringer in #1465
• fix docs for mps #1455 by @carsen-stringer in #1466
• fix: avoid saving empty .zip archive by @camlloyd in #1242
• quick fix for the raised issue #12208 regarding the functions of io.imsave() and io.save_rois() by @Jiadalee in #1211
• fixed "RuntimeError: structure and input must have equal rank" by @mese79 in #1123
• small fixes for uint32, colab notebooks, cellprob return by @carsen-stringer in #1467
• Fix: Exit early from update_orthos() when orthos are off by @mrariden in #1458
• update doc string #1469 and import error message by @carsen-stringer in #1470
• Slice 3d labels by @carsen-stringer in #1472

New Contributors

• @camlloyd made their first contribution in #1242
• @Jiadalee made their first contribution in #1211
• @mese79 made their first contribution in #1123

Full Changelog: v4.1.1...v4.2.1.1

v4.1.1

What's Changed

• Gray saturation fix: drop blank channels from mean by @mrariden in #1423

Full Changelog: v4.1.0...v4.1.1

v4.1.0

What's Changed

• add support for anisotropic values of the flow3D_smooth by @yuriyzubov in #1408
• Fix run dynamics and flow docs by @mrariden in #1416
• Refactor eval by @mrariden in #1417

New Contributors

• @yuriyzubov made their first contribution in #1408

Full Changelog: v4.0.9...v4.1.0

v4.0.9

What's Changed

• Revert to training in float32 by @mrariden in #1390
• Update return tuple in 'eval' function documentation by @ralfox in #1403
• Set transforms logging to debug level and add logger setup documentation by @mrariden in #1412
• Fix GH workflows tests: configure tox to pass environment variables for PyTorch by @qin-yu in #1411
• guess Z dim is 0 instead of argmin of dims by @mrariden in #1409
• .nd2 files now load correctly in the gui by @tlambert03 in #1406

New Contributors

• @ralfox made their first contribution in #1403
• @tlambert03 made their first contribution in #1406

Full Changelog: v4.0.8...v4.0.9

v4.0.8

What's Changed

• update default training lr to 1e-5, and update docs by @mrariden in #1349
• allow relative savedir path by @mrariden in #1354
• Allow specify gpu device number by @mrariden in #1311
• fail on loading cp3 model in cp4 by @mrariden in #1368
• fix(utils): fixes a bug in fill_holes_and_remove_small_masks by @rubencardenes in #1367
• Fix ImageJ ROIs with inscrutable names by @MarinManuel in #1358
• Batch img refactor (allow batch input of images) by @mrariden in #1374

New Contributors

• @rubencardenes made their first contribution in #1367
• @MarinManuel made their first contribution in #1358

Full Changelog: v4.0.7...v4.0.8

v4.0.7

What's Changed

• fix Overflow error in saving flows in io.save_masks by @pywugate in #1261
• Patch training bfloat16 on MPS bug by @mrariden in #1278
• Allows to open external links by @KacperPietka in #1265
• Change style vector to return zeros instead of random numbers and update docs. Remove CP4 welcome message. @mrariden in #1289 and #1298
• Fix typo in deprecation warning message by @cgohlke in #1322
• Increase block size from 224 to 256 in eval method to make consistent with documentation by @MassEast in #1330
• more verbose tox/pytest tests by @mrariden in #1332
• Faq update by @mrariden in #1341

New Contributors

• @pywugate made their first contribution in #1261
• @KacperPietka made their first contribution in #1265
• @MassEast made their first contribution in #1330

Full Changelog: v4.0.6...v4.0.7

v4.0.6

What's Changed

• Fix: Add missing _label_overlap function and integrate into metrics by @mario-koddenbrock in #1246
• scale niter based on image scaling by @mrariden in #1259
• Fix 3d reading cli by @mrariden in #1258
• add notebook for running cellpose3 with denoising and segmentation by @mrariden in #1260

New Contributors

• @mario-koddenbrock made their first contribution in #1246

Full Changelog: v4.0.5...v4.0.6

v4.0.5

Updates to Cellpose-SAM

• resample added back to CLI for diameter changes. Note that the diameter parameter is only used for image rescaling and is not always necessary to provide in CPSAM compared to CP3.
• default model weights are changed to torch.bfloat16 instead of previous 32-bit float data type. This decreases the model size by ~50% and speeds up runtime by ~40%, therefore helping with OOM issues. Changing back to the 32-bit model can be done during CellposeModel instantiation by setting use_bfloat16=False.
• Improved 3d image dimension guessing
• Fixed a bug that ran the network 4x more often than necessary when image size was <256 length or width.
• Added some documentation
• Updated notebooks
• Some more comprehensive testing

Changelog: v4.0.4...v4.0.5

v4.0.4

Introducing Cellpose-SAM!

New in cellpose v4

• Cellpose-SAM aka CP4 is fully integrated in to the CLI/API/GUI and notebooks
• cellpose.models.Cellpose class was removed, users should now only use cellpose.models.CellposeModel
• Size estimation via cellpose.models.SizeModel class was removed. Users don't have to use the diameter setting since CP4 is much more robust to object size.
• Denoising components via cellpose.denoise is currently not available.
• Added more comprehensive testing
• Better utilization of MPS for Mac users when doing 3D segmentation
• Added --save_each flag for training from the CLI
• More details are available in the paper, talk, and in the docs.

New Contributors

• @KevinCortacero made their first contribution in #1080
• @unikill066 made their first contribution in #1159

Changelog for early cellpose v4 changes: v4.0.1...v4.0.4

v3.1.1.2

This is the last release of Cellpose3 before the Cellpose-SAM release.

What's Changed

• fix tiling bug when augment=True
• Documented using masks and preprocessing_steps in distributed rst by @GFleishman in #1108
• Compress TIFFs on save by @qin-yu in #1131
• Report errors when saving masks to *_seg.npy by @deroulers in #1129
• Correct GUI initialization order of load_3D for modular usage. by @biswajitSM in #1119
• Optimize fill_holes_and_remove_small_masks by @Tomvl117 in #1116
• pyarrow missing by @GFleishman in #1109

New Contributors

• @biswajitSM made their first contribution in #1119
• @Tomvl117 made their first contribution in #1116

Full Changelog: v3.1.1.1...v3.1.1.2