dsa-helpers 0.5.dev0__tar.gz → 0.6__tar.gz

{dsa_helpers-0.5.dev0 → dsa_helpers-0.6}/.gitignore

@@ -8,4 +8,5 @@ dist/
 .env
 
 # Ignore data directory created for testing the package.
-tests/data/
+tests/data/
+tests/models/

{dsa_helpers-0.5.dev0/dsa_helpers.egg-info → dsa_helpers-0.6}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: dsa_helpers
-Version: 0.5.dev0
+Version: 0.6
 Summary: Utility functions for working with the DSA girder client.
 Author-email: "Juan C. Vizcarra" <jvizcar@emory.edu>
 Maintainer-email: "Juan C. Vizcarra" <jvizcar@emory.edu>
@@ -8,7 +8,7 @@ Project-URL: Repository, https://github.com/Gutman-Lab/dsa-python-utils
 Keywords: dsa,girder,girder_client
 Classifier: Programming Language :: Python :: 3
 Classifier: License :: OSI Approved :: MIT License
-Requires-Python: >=3.9
+Requires-Python: >=3.8
 Description-Content-Type: text/markdown
 License-File: LICENSE
 Requires-Dist: girder-client
@@ -22,9 +22,9 @@ Requires-Dist: dash
 Requires-Dist: dash_bootstrap_components
 Requires-Dist: shapely
 Requires-Dist: geopandas
+Requires-Dist: transformers[torch]
 Requires-Dist: torch
 Requires-Dist: datasets
-Requires-Dist: transformers
 Requires-Dist: torchvision
 
 # Digital Slide Archive (DSA) Helpers

{dsa_helpers-0.5.dev0 → dsa_helpers-0.6}/dsa_helpers/image_utils.py

@@ -14,6 +14,7 @@ def tile_image(
     fill: int | tuple = (255, 255, 255),
     prepend_name: str = "",
     overwrite: bool = False,
+    grayscale: bool = False,
 ) -> pd.DataFrame:
     """Tile an image into smaller images.
 
@@ -28,6 +29,8 @@ def tile_image(
         prepend_name (str, optional): A string to prepend to the tile names.
         overwrite (bool, optional): Whether to overwrite existing images.
             Defaults to False.
+        grayscale (bool, optional): Whether to save images as grayscale instead
+            of the default RGB. Defaults to False.
 
     Returns:
         pandas.DataFrame: A DataFrame with the tile locations.
@@ -50,18 +53,21 @@ def tile_image(
 
     xys = [(x, y) for x in range(0, w, stride) for y in range(0, h, stride)]
 
+    if prepend_name:
+        prepend_name += "_"
+
     for xy in xys:
         # Get the top left and bottom right coordinates of the tile.
         x, y = xy
 
         # Filepath for saving tile.
-        save_fp = save_loc / f"{prepend_name}_x{x}y{y}tilesize{tile_size}.png"
+        save_fp = save_loc / f"{prepend_name}x{x}y{y}tilesize{tile_size}.png"
 
         if not save_fp.is_file() or overwrite:
             # Get the tile from the image.
             tile_img = img[y : y + tile_size, x : x + tile_size]
 
-            imwrite(save_fp, tile_img)
+            imwrite(save_fp, tile_img, grayscale=grayscale)
 
         df_data.append([save_fp, x, y, tile_size])
 

dsa_helpers-0.6/dsa_helpers/ml/__init__.py

@@ -0,0 +1,17 @@
+# Allow modules to be imported.
+from . import (
+    object_detection,
+    datasets,
+    metrics,
+    transforms,
+    segformer_semantic_segmentation,
+)
+
+# Modules that are imported from * notation.
+__all__ = [
+    "object_detection",
+    "datasets",
+    "metrics",
+    "transforms",
+    "segformer_semantic_segmentation",
+]

dsa_helpers-0.6/dsa_helpers/ml/segformer_semantic_segmentation/__init__.py

@@ -0,0 +1,3 @@
+from .inference import inference
+
+__all__ = ["inference"]

dsa_helpers-0.6/dsa_helpers/ml/segformer_semantic_segmentation/inference.py

@@ -0,0 +1,121 @@
+import numpy as np
+from pathlib import Path
+import cv2 as cv
+import shutil
+
+import transformers
+from transformers import SegformerImageProcessor
+
+import torch
+import torch.nn as nn
+
+from ... import imread
+from ...image_utils import tile_image
+
+
+def inference(
+    img: np.ndarray,
+    model: transformers.models.segformer.modeling_segformer.SegformerForSemanticSegmentation,
+    device: str | None = None,
+    tile_size: int = 512,
+    temp_dir: str = ".temp",
+    batch_size: int = 8,
+) -> tuple[np.ndarray, list[np.ndarray]]:
+    """Inference on an image using SegFormer model for semantic segmentation.
+
+    Args:
+        img (numpy.ndarray): Image to segment.
+        model (transformers.models.segformer.modeling_segformer.SegformerForSemanticSegmentation):
+            Pre-trained model.
+        device (str | None): Device to use. Defaults to None, in which case
+            it is inferred based on devices available.
+        tile_size (int): Size of the tiles. Defaults to 512.
+        temp_dir (str): Temporary directory to save tiles. Defaults to ".temp".
+        batch_size (int): Batch size. Defaults to 8.
+
+    Returns:
+        tuple[numpy.ndarray, list[numpy.ndarray]]: The predicted label mask and
+            the contours of the mask.
+
+    """
+    # Create location to save tile images.
+    tile_dir = Path(temp_dir)
+    tile_dir.mkdir(exist_ok=True, parents=True)
+
+    # Tile the image.
+    tiles_df = tile_image(
+        img,
+        str(tile_dir),
+        tile_size=tile_size,
+    )
+
+    # Create instance of the image processor.
+    img_processor = SegformerImageProcessor()
+
+    if device is None:
+        device = "cuda" if torch.cuda.is_available() else "cpu"
+
+    print("Using device:", device)
+
+    # Set model to eval mode.
+    model.to(device)  # to device
+    model.eval()
+
+    # Create an empty mask the size of the image.
+    h, w = img.shape[:2]
+    predicted_mask = np.zeros((h + tile_size, w + tile_size), dtype=np.uint8)
+
+    # Loop through tile images in batch size.
+    for i in range(0, len(tiles_df), batch_size):
+        # Get batch of tile images.
+        batch = tiles_df.iloc[i : i + batch_size]
+
+        # Read all the images in the batch.
+        images = [imread(fp) for fp in batch["fp"]]
+
+        # Process all the images.
+        inputs = img_processor(
+            images=images,
+            return_tensors="pt",
+        )
+
+        # Send to the same device as model.
+        inputs = inputs.to(device)
+
+        with torch.no_grad():
+            # Sent inputs through the model.
+            outputs = model(**inputs)
+
+            # Get out the logits.
+            logits = outputs.logits
+
+            # Reshape the logits to the original image size.
+            upsampled_logits = nn.functional.interpolate(
+                logits,
+                size=(tile_size, tile_size),  # (height, width)
+                mode="bilinear",
+                align_corners=False,
+            )
+
+            # Take the max value of the class dimension.
+            pred_seg = upsampled_logits.argmax(dim=1)
+
+            # Place tile predictions into the overall predicted mask.
+            for j in range(len(batch)):
+                r = batch.iloc[j]
+                x, y = r["x"], r["y"]
+                predicted_mask[y : y + tile_size, x : x + tile_size] = (
+                    pred_seg[j].cpu().numpy()
+                ).astype(np.uint8)
+
+    # Remove the temp directory.
+    shutil.rmtree(temp_dir)
+
+    # Remove edges.
+    predicted_mask = predicted_mask[:h, :w]
+
+    # Extract the contours of the mask.
+    contours = cv.findContours(predicted_mask, cv.RETR_TREE, cv.CHAIN_APPROX_SIMPLE)[0]
+
+    # Return the mask.
+    return predicted_mask, contours

{dsa_helpers-0.5.dev0 → dsa_helpers-0.6}/dsa_helpers/mongo_utils.py

@@ -9,6 +9,7 @@ from typing import Union
 
 
 def get_mongo_client(
+    protocol: str | None = None,
     host: str | None = None,
     username: str | None = None,
     password: str | None = None,
@@ -20,6 +21,9 @@ def get_mongo_client(
     not present it sets them to default values.
 
     Args:
+        protocol (str, optional): The protocol of the mongo server. If None it
+            look for the environment variable "MONGO_PROTOCOL". If this is
+            missing it will default to "mongodb".
         host (str, optional): The host name of the mongo server. If None it will
             look for the environment variable "MONGO_HOST_NAME". If this is
             missing it will default to "mongodb".
@@ -39,6 +43,8 @@ def get_mongo_client(
         The mongo client.
 
     """
+    if protocol is None:
+        protocol = getenv("MONGO_PROTOCOL", "mongodb")
     if username is None:
         username = getenv("MONGO_INITDB_ROOT_USERNAME", "docker")
     if password is None:
@@ -48,7 +54,7 @@ def get_mongo_client(
     if port is None:
         port = int(getenv("MONGO_HOST_PORT", 27017))
 
-    return pymongo.MongoClient(f"mongodb://{username}:{password}@{host}:{port}")
+    return pymongo.MongoClient(f"{protocol}://{username}:{password}@{host}:{port}")
 
 
 def chunks(lst: list, n=500):

{dsa_helpers-0.5.dev0 → dsa_helpers-0.6/dsa_helpers.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: dsa_helpers
-Version: 0.5.dev0
+Version: 0.6
 Summary: Utility functions for working with the DSA girder client.
 Author-email: "Juan C. Vizcarra" <jvizcar@emory.edu>
 Maintainer-email: "Juan C. Vizcarra" <jvizcar@emory.edu>
@@ -8,7 +8,7 @@ Project-URL: Repository, https://github.com/Gutman-Lab/dsa-python-utils
 Keywords: dsa,girder,girder_client
 Classifier: Programming Language :: Python :: 3
 Classifier: License :: OSI Approved :: MIT License
-Requires-Python: >=3.9
+Requires-Python: >=3.8
 Description-Content-Type: text/markdown
 License-File: LICENSE
 Requires-Dist: girder-client
@@ -22,9 +22,9 @@ Requires-Dist: dash
 Requires-Dist: dash_bootstrap_components
 Requires-Dist: shapely
 Requires-Dist: geopandas
+Requires-Dist: transformers[torch]
 Requires-Dist: torch
 Requires-Dist: datasets
-Requires-Dist: transformers
 Requires-Dist: torchvision
 
 # Digital Slide Archive (DSA) Helpers

{dsa_helpers-0.5.dev0 → dsa_helpers-0.6}/dsa_helpers.egg-info/SOURCES.txt

@@ -30,6 +30,8 @@ dsa_helpers/ml/object_detection/__init__.py
 dsa_helpers/ml/object_detection/get_pearce_roi_images.py
 dsa_helpers/ml/object_detection/tile_image.py
 dsa_helpers/ml/object_detection/utils.py
+dsa_helpers/ml/segformer_semantic_segmentation/__init__.py
+dsa_helpers/ml/segformer_semantic_segmentation/inference.py
 dsa_helpers/ml/transforms/__init__.py
 dsa_helpers/ml/transforms/segformer_transforms.py
 tests/docker-compose.yaml

{dsa_helpers-0.5.dev0 → dsa_helpers-0.6}/dsa_helpers.egg-info/requires.txt

@@ -9,7 +9,7 @@ dash
 dash_bootstrap_components
 shapely
 geopandas
+transformers[torch]
 torch
 datasets
-transformers
 torchvision

{dsa_helpers-0.5.dev0 → dsa_helpers-0.6}/pyproject.toml

@@ -15,9 +15,9 @@ dependencies = [
     'dash_bootstrap_components',
     'shapely',
     'geopandas',
+    "transformers[torch]",
     "torch",
     "datasets",
-    "transformers",
     "torchvision"
 ]
 name = "dsa_helpers"
@@ -29,13 +29,13 @@ maintainers = [
 ]
 description = "Utility functions for working with the DSA girder client."
 readme = "README.md"
-requires-python = ">=3.9"
+requires-python = ">=3.8"
 classifiers = [
     "Programming Language :: Python :: 3",
     "License :: OSI Approved :: MIT License",
 ]
 keywords = ["dsa", "girder", "girder_client"]
-version = "v0.5dev0"
+version = "v0.6"
 
 [project.urls]
 Repository = "https://github.com/Gutman-Lab/dsa-python-utils"

{dsa_helpers-0.5.dev0 → dsa_helpers-0.6}/requirements.txt

@@ -2,18 +2,22 @@
 build
 twine
 
+# For tests.
+matplotlib
+jupyter
+
+# Mirrored requirements from pyproject.toml
 girder-client
 numpy
 setuptools
 opencv-python
 pandas
 pymongo
-jupyter
 pillow
 dash
 dash_bootstrap_components
 python-dotenv
+transformers[torch]
 torch
 datasets
-transformers
 torchvision