Merge pull request #516 from computational-cell-analytics/dev

Merge dev in preparation for new release

.github/workflows/build_installers.yaml

@@ -1,9 +1,7 @@
 name: build_installers
 
 on:
-  push:
-    tags:
-      - '**'
+  workflow_dispatch:
 
 jobs:
   build_intaller:
@@ -13,15 +11,20 @@ jobs:
     strategy:
       fail-fast: false
       matrix:
-        os: [windows-latest, ubuntu-latest, macos-latest]
+        os: [windows-latest, ubuntu-latest]  # macos-latest
 
     env:
       PREPARE_SCRIPT: |
         cd deployment
-        conda install -y -c conda-forge constructor
-        conda install -y -c conda-forge ruamel.yaml
-        conda install -y -c conda-forge mamba
-        mamba env create --file=env.yaml
+        mamba install -y -c conda-forge constructor
+        mamba install -y -c conda-forge ruamel.yaml
+
+        mamba env create --file=../environment_cpu.yaml -n __MICROSAM_BUILD_ENV__
+        mamba activate __MICROSAM_BUILD_ENV__
+        # TODO get the current version here and use it for pinning or enable passing this from dispatch
+        mamba install -c conda-forge micro_sam
+        mamba activate base
+
       RUN_SCRIPT: |
         python version_getter.py
         mkdir ./${{ matrix.os }}_x86_64
@@ -30,44 +33,47 @@ jobs:
     steps:
       - name: checkout
         uses: actions/checkout@v4
+        with:
+          ref: ${{ github.ref }}
 
       - name: setup conda
-        if: matrix.os == 'windows-latest' || matrix.os == 'ubuntu-latest'
+        if: matrix.os == "windows-latest" || matrix.os == "ubuntu-latest"
         uses: conda-incubator/setup-miniconda@v2
         with:
           miniconda-version: "latest"
           auto-activate-base: true
           activate-environment: ""
+          mamba-version: "*"
         env:
           ACTIONS_ALLOW_UNSECURE_COMMANDS: true
 
       - name: build ${{ matrix.os }}_x86_64
-        if: matrix.os == 'windows-latest'
+        if: matrix.os == "windows-latest"
         shell: pwsh
         run: |
           ${{ env.PREPARE_SCRIPT }}
-          conda activate __MICROSAM_BUILD_ENV__
-          conda install -c anaconda menuinst
+          mamba activate sam
+          mamba install -c anaconda menuinst
           python windows_menu_setup.py
           conda activate base
           ${{ env.RUN_SCRIPT }}
 
       - name: build ${{ matrix.os }}_x86_64
-        if: matrix.os == 'ubuntu-latest'
+        if: matrix.os == "ubuntu-latest"
         shell: bash -el {0}
         run: |
           ${{ env.PREPARE_SCRIPT }}
           ${{ env.RUN_SCRIPT }}
 
       - name: build ${{ matrix.os }}_x86_64_step1
-        if: matrix.os == 'macos-latest'
+        if: matrix.os == "macos-latest"
         shell: bash -el {0}
         run: |
           brew install micromamba
           /usr/local/opt/micromamba/bin/micromamba shell init -s bash -p ~/micromamba
 
       - name: build ${{ matrix.os }}_x86_64_step2
-        if: matrix.os == 'macos-latest'
+        if: matrix.os == "macos-latest"
         shell: bash -el {0}
         run: |
           cd deployment

.gitignore

@@ -173,3 +173,7 @@ cython_debug/
 #  and can be added to the global gitignore or merged into this file.  For a more nuclear
 #  option (not recommended) you can uncomment the following to ignore the entire idea folder.
 #.idea/
+
+# Torch-em stuff
+checkpoints/
+logs/

README.md

@@ -3,9 +3,11 @@
 [![codecov](https://codecov.io/gh/computational-cell-analytics/micro-sam/graph/badge.svg?token=7ETPP5CABP)](https://codecov.io/gh/computational-cell-analytics/micro-sam)
 [![DOI](https://zenodo.org/badge/DOI/10.5281/zenodo.7919746.svg)](https://doi.org/10.5281/zenodo.7919746)
 
-# SegmentAnything for Microscopy
+# Segment Anything for Microscopy
 
-Tools for segmentation and tracking in microscopy build on top of [SegmentAnything](https://segment-anything.com/).
+**Attention: We are currently updating our software to a new release that will improve it and introduce new features. The documentation is not up-to-date with these changes yet, we will update it as soon as possible!**
+
+Tools for segmentation and tracking in microscopy build on top of [Segment Anything](https://segment-anything.com/).
 Segment and track objects in microscopy images interactively with a few clicks!
 
 We implement napari applications for:
@@ -22,21 +24,16 @@ If you run into any problems or have questions regarding our tool please open an
 
 ## Installation and Usage
 
-You can install `micro_sam` via conda:
-```
-conda install -c conda-forge micro_sam napari pyqt
-```
-You can then start the `micro_sam` tools by running `$ micro_sam.annotator` in the command line.
+Please check [the documentation](https://computational-cell-analytics.github.io/micro-sam/) for details on how to install and use `micro_sam`. You can also find a quickstart guide in [this video](TODO) and find all video tutorials [here](https://www.youtube.com/watch?v=ket7bDUP9tI&list=PLwYZXQJ3f36GQPpKCrSbHjGiH39X4XjSO&pp=gAQBiAQB). 
 
-For an introduction in how to use the napari based annotation tools check out [the video tutorials](https://www.youtube.com/watch?v=ket7bDUP9tI&list=PLwYZXQJ3f36GQPpKCrSbHjGiH39X4XjSO&pp=gAQBiAQB).
-Please check out [the documentation](https://computational-cell-analytics.github.io/micro-sam/) for more details on the installation and usage of `micro_sam`.
 
 ## Contributing
 
 We welcome new contributions!
 
 If you are interested in contributing to micro-sam, please see the [contributing guide](doc/contributing.md) and [developer documentation](doc/development.md). The first step is to [discuss your idea in a new issue](https://github.com/computational-cell-analytics/micro-sam/issues/new) with the current developers.
 
+
 ## Citation
 
 If you are using this repository in your research please cite

deployment/construct.yaml

@@ -1,13 +1,13 @@
-name: micro_sam
-version: 0.1.0
-license_file: ../LICENSE
-installer_type: pkg  #[osx]  # This will trigger pkg build on Mac Os. On windows and linux, native build will be done and this has no effect.
-environment: __MICROSAM_BUILD_ENV__
-welcome_image: ../doc/images/micro-sam-logo.png
-header_image: ../doc/images/micro-sam-logo.png
-icon_image: ../doc/images/micro-sam-logo.png
-channels:
-- conda-forge
-welcome_text: Install Segment Anything for Microscopy.
-conclusion_text: Segment Anything for Microscopy has been installed.
-initialize_by_default: false
+name: micro_sam
+version: 0.0.1
+license_file: ../LICENSE
+installer_type: pkg  #[osx]  # This will trigger pkg build on Mac Os. On windows and linux, native build will be done and this has no effect.
+environment: __MICROSAM_BUILD_ENV__
+welcome_image: ../doc/images/micro-sam-logo.png
+header_image: ../doc/images/micro-sam-logo.png
+icon_image: ../doc/images/micro-sam-logo.png
+channels:
+  - conda-forge
+welcome_text: Install Segment Anything for Microscopy.
+conclusion_text: Segment Anything for Microscopy has been installed.
+initialize_by_default: false

deployment/version_getter.py

@@ -6,10 +6,10 @@
 yaml.preserve_quotes = True
 ctor_fname = os.path.join("construct.yaml")
 
-with open(ctor_fname, 'r') as stream:
+with open(ctor_fname, "r") as stream:
     ctor_conf = yaml.load(stream)
 
 ctor_conf["version"] = runpy.run_path(os.path.join("..", "micro_sam", "__version__.py"))["__version__"]
 
-with open(ctor_fname, 'w') as outfile:
+with open(ctor_fname, "w") as outfile:
     yaml.dump(ctor_conf, outfile)

deployment/windows_menu.json

@@ -3,7 +3,7 @@
     "menu_items":
         [
             {
-                "script": "${PREFIX}/Scripts/micro_sam.annotator.exe",
+                "script": "${PREFIX}/Scripts/napari.exe",
 				"name": "micro_sam",
 				"icon": "${PREFIX}/Menu/micro-sam-logo.ico",
 				"workdir": "${PREFIX}",

development/annotators_in_image_series.py

@@ -0,0 +1,103 @@
+import os
+
+import h5py
+from math import ceil
+
+from micro_sam.sam_annotator import image_series_annotator, annotator_3d
+
+
+DATA_ROOT = "/home/anwai/data/lucchi/lucchi_test.h5"
+EMBEDDING_ROOT = "/home/anwai/embeddiings/test/"
+OUTPUT_ROOT = "/home/anwai/data/lucchi/outputs/"
+
+
+def _get_volume(volume_path):
+    """Getting the lucchi test volume"""
+    with h5py.File(volume_path, "r") as f:
+        raw = f["raw"][:]
+        labels = f["labels"][:]
+
+    return raw, labels
+
+
+def segment_volume(input_volume, embedding_path):
+    """Load the entire volume in the tool for segmentation.
+    """
+    assert input_volume.ndim == 3
+    annotator_3d(
+        image=input_volume,
+        embedding_path=embedding_path,
+        model_type="vit_b_em_organelles",
+        tile_shape=None,
+        halo=None,
+    )
+
+
+def segment_each_slice(input_volume, embedding_dir, output_folder):
+    """Load each slice from the volume one-by-one in the tool for segmentation.
+    """
+    assert input_volume.ndim == 3
+
+    all_slices = [each_slice for each_slice in input_volume]
+    image_series_annotator(
+        images=all_slices,
+        output_folder=output_folder,
+        model_type="vit_b_em_organelles",
+        embedding_path=embedding_dir,
+        tile_shape=None,
+        halo=None,
+    )
+
+
+def segment_each_n_slices(z_batch, input_volume, embedding_dir, output_folder):
+    """Load n slices from the volume one-by-one in the tool for segmentation.
+    """
+    assert input_volume.ndim == 3
+
+    n_z_slices = input_volume.shape[0]
+    all_z_idxx = int(ceil(n_z_slices / z_batch))
+
+    all_per_n_slices_volumes = []
+    for z_id in range(all_z_idxx):
+        z_start = z_id * z_batch
+        z_stop = min((z_id + 1) * z_batch, n_z_slices)
+
+        batch_volume = input_volume[z_start: z_stop]
+        all_per_n_slices_volumes.append(batch_volume)
+
+    print(f"We split the volume into {len(all_per_n_slices_volumes)} sub-volumes.")
+    image_series_annotator(
+        images=all_per_n_slices_volumes,
+        output_folder=output_folder,
+        model_type="vit_b_em_organelles",
+        embedding_path=embedding_dir,
+        tile_shape=None,
+        halo=None,
+        is_volumetric=True,
+    )
+
+
+def main():
+    raw, _ = _get_volume(DATA_ROOT)
+
+    # segment_volume(
+    #     input_volume=raw,
+    #     embedding_path=os.path.join(EMBEDDING_ROOT, "lucchi_3d_volume")
+    # )
+
+    # segment_each_slice(
+    #     input_volume=raw,
+    #     embedding_dir=os.path.join(EMBEDDING_ROOT, "lucchi_2d_per_slice"),
+    #     output_folder=os.path.join(OUTPUT_ROOT, "per_slice_segmentation")
+    # )
+
+    segment_each_n_slices(
+        z_batch=15,
+        input_volume=raw,
+        embedding_dir=os.path.join(EMBEDDING_ROOT, "lucchi_3d_per_n_slices"),
+        output_folder=os.path.join(OUTPUT_ROOT, "per_n_slices")
+    )
+
+
+if __name__ == "__main__":
+    main()

doc/bioimageio/em_organelles_v2.md

@@ -0,0 +1,17 @@
+# Segment Anything for Electron Microscopy
+
+This is a [Segment Anything](https://segment-anything.com/) model that was specialized for segmenting mitochondria and nuclei in electron microscopy with [micro_sam](https://github.com/computational-cell-analytics/micro-sam).
+This model uses a %s vision transformer as image encoder.
+
+Segment Anything is a model for interactive and automatic instance segmentation.
+We improve it for light microscopy by finetuning on a large and diverse microscopy dataset.
+It should perform well for segmenting mitochondria and nuclei in electron microscopy. It can also work well for other organelles, but was not explicitly trained for this purpose. You may get better results for other organelles (e.g. ER or Golgi) with the default Segment Anything models.
+
+See [the dataset overview](https://github.com/computational-cell-analytics/micro-sam/blob/master/doc/datasets/em_organelles_v%i.md) for further informations on the training data and the [micro_sam documentation](https://computational-cell-analytics.github.io/micro-sam/micro_sam.html) for details on how to use the model for interactive and automatic segmentation.
+
+
+## Validation
+
+The easiest way to validate the model is to visually check the segmentation quality for your data.
+If you have annotations you can use for validation you can also quantitative validation, see [here for details](https://github.com/computational-cell-analytics/micro-sam/blob/master/doc/bioimageio/validation.md).
+Please note that the required quality for segmentation always depends on the analysis task you want to solve.