Skip to content

Commit

Permalink
add vera's comments and tqdm
Browse files Browse the repository at this point in the history
  • Loading branch information
@ahillsley
ahillsley committed May 8, 2024
1 parent efde7c5 commit c706cc8
Showing 1 changed file with 47 additions and 30 deletions.
77 changes: 47 additions & 30 deletions 3-instance_solution.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -84,8 +84,9 @@ def compute_sdt(labels: np.ndarray, scale: int):
# compute the distance transform inside of each individual object and the background

# create the signed distance transform
# remember that it should be positive inside the objects and negative outside

# scale the distances so that they are between -1 and 1 (hint: np.tanh)
# scale the distances with the scale parameter and then normalize so that they are between -1 and 1 (hint: np.tanh)

# be sure to return your solution as type 'float'
return
Expand Down Expand Up @@ -125,21 +126,25 @@ def compute_sdt(labels: np.ndarray, scale: int = 5):
# %%
# Visualize the signed distance transform using the function you wrote above.
root_dir = "/group/dl4miacourse/segmentation/nuclei_train_data" # the directory with all the training samples
samples = os.listdir(root_dir)
samples = os.listdir(root_dir)
idx = np.random.randint(len(samples)) # take a random sample.
img = tifffile.imread(os.path.join(root_dir, samples[idx], 'image.tif')) # get the image
label = tifffile.imread(os.path.join(root_dir, samples[idx], 'label.tif')) # get the image
img = tifffile.imread(
os.path.join(root_dir, samples[idx], "image.tif")
) # get the image
label = tifffile.imread(
os.path.join(root_dir, samples[idx], "label.tif")
) # get the image
sdt = compute_sdt(label)
plot_two(img, sdt, label="SDT")


# %% tags [markdown]
# <b>Questions</b>:
# 1. _Why do we need to normalize the distances between -1 and 1_?
#
#
#
# 2. _What is the effect of changing the scale value? What do you think is a good default value_?
#
#

# %% [markdown] tags=["solution"]
# <b>Questions</b>:
Expand All @@ -164,7 +169,9 @@ class SDTDataset(Dataset):
"""A PyTorch dataset to load cell images and nuclei masks."""

def __init__(self, root_dir, transform=None, img_transform=None, return_mask=False):
self.root_dir = "/group/dl4miacourse/segmentation/" + root_dir # the directory with all the training samples
self.root_dir = (
"/group/dl4miacourse/segmentation/" + root_dir
) # the directory with all the training samples
self.samples = os.listdir(self.root_dir) # list the samples
self.return_mask = return_mask
self.transform = (
Expand All @@ -182,7 +189,7 @@ def __init__(self, root_dir, transform=None, img_transform=None, return_mask=Fal

self.loaded_imgs = [None] * len(self.samples)
self.loaded_masks = [None] * len(self.samples)
for sample_ind in range(len(self.samples)):
for sample_ind in tqdm(range(len(self.samples))):
img_path = os.path.join(
self.root_dir, self.samples[sample_ind], "image.tif"
)
Expand All @@ -195,7 +202,7 @@ def __init__(self, root_dir, transform=None, img_transform=None, return_mask=Fal
mask = Image.open(mask_path)
mask.load()
self.loaded_masks[sample_ind] = mask

# get the total number of samples
def __len__(self):
return len(self.samples)
Expand All @@ -204,6 +211,7 @@ def __len__(self):
def __getitem__(self, idx):

# TODO: Modify this function to return an image and sdt pair
# add a call to the create sdt_target function somewhere below ... where is the challenge

# we'll be using the Pillow library for reading files
# since many torchvision transforms operate on PIL images
Expand All @@ -226,9 +234,10 @@ def __getitem__(self, idx):
# only need the image and sdt for training
return image, sdt

def create_sdt_target(self):
def create_sdt_target(self, mask):
# TODO: Fill in function
# make sure this function is returning a torch tensor
# chose a scale value to use here, it's fine to hard code it in

return

Expand All @@ -238,7 +247,9 @@ class SDTDataset(Dataset):
"""A PyTorch dataset to load cell images and nuclei masks."""

def __init__(self, root_dir, transform=None, img_transform=None, return_mask=False):
self.root_dir = "/group/dl4miacourse/segmentation/" + root_dir # the directory with all the training samples
self.root_dir = (
"/group/dl4miacourse/segmentation/" + root_dir
) # the directory with all the training samples
self.samples = os.listdir(self.root_dir) # list the samples
self.return_mask = return_mask
self.transform = (
Expand All @@ -256,7 +267,7 @@ def __init__(self, root_dir, transform=None, img_transform=None, return_mask=Fal

self.loaded_imgs = [None] * len(self.samples)
self.loaded_masks = [None] * len(self.samples)
for sample_ind in range(len(self.samples)):
for sample_ind in tqdm(range(len(self.samples))):
img_path = os.path.join(
self.root_dir, self.samples[sample_ind], "image.tif"
)
Expand Down Expand Up @@ -302,6 +313,7 @@ def create_sdt_target(self, mask):
sdt_target = transforms.ToTensor()(sdt_target_array)
return sdt_target.float()


# %% [markdown]
# ### Test your function
#
Expand Down Expand Up @@ -343,13 +355,13 @@ def create_sdt_target(self, mask):
out_channels=...,
)

learning_rate=1e-4
learning_rate = 1e-4

# Choose a loss function.

# Choose an optimizer.

# Use the train function provided in local.py
# Use the train function provided in local.py
# to train the model for 20 epochs.

for epoch in range(20):
Expand Down Expand Up @@ -448,6 +460,7 @@ def find_local_maxima(distance_transform, min_dist_between_points):
# %% tags=["solution"]
from scipy.ndimage import label, maximum_filter


def find_local_maxima(distance_transform, min_dist_between_points):
# Use `maximum_filter` to perform a maximum filter convolution on the distance_transform
max_filtered = maximum_filter(distance_transform, min_dist_between_points)
Expand All @@ -457,16 +470,19 @@ def find_local_maxima(distance_transform, min_dist_between_points):

return seeds, n


# %%
# test your function.
from local import test_maximum

test_maximum(find_local_maxima)

# %% [markdown]
# We now use this function to find the seeds for the watershed.
# %%
from skimage.segmentation import watershed


def watershed_from_boundary_distance(
boundary_distances: np.ndarray,
inner_mask: np.ndarray,
Expand Down Expand Up @@ -646,11 +662,14 @@ def get_inner_mask(pred, threshold):
# create a new dataset for affinities
from local import compute_affinities


class AffinityDataset(Dataset):
"""A PyTorch dataset to load cell images and nuclei masks"""

def __init__(self, root_dir, transform=None, img_transform=None, return_mask=False):
self.root_dir = "/group/dl4miacourse/segmentation/" + root_dir # the directory with all the training samples
self.root_dir = (
"/group/dl4miacourse/segmentation/" + root_dir
) # the directory with all the training samples
self.samples = os.listdir(self.root_dir) # list the samples
self.return_mask = return_mask
self.transform = (
Expand All @@ -668,7 +687,7 @@ def __init__(self, root_dir, transform=None, img_transform=None, return_mask=Fal

self.loaded_imgs = [None] * len(self.samples)
self.loaded_masks = [None] * len(self.samples)
for sample_ind in range(len(self.samples)):
for sample_ind in tqdm(range(len(self.samples))):
img_path = os.path.join(
self.root_dir, self.samples[sample_ind], "image.tif"
)
Expand Down Expand Up @@ -735,22 +754,21 @@ def create_aff_target(self, mask):
# %%
# define the model
unet = UNet(
depth= ...,
depth=...,
in_channels=...,
downsample_factor=...,
final_activation=...,
final_activation=...,
out_channels=...,
)

learning_rate = 1e-4
# specify loss
# specify loss

# specify optimizer

# add training loop



# %% tags=["solution"]
unet = UNet(
depth=4,
Expand Down Expand Up @@ -819,23 +837,23 @@ def create_aff_target(self, mask):
)
for idx, (image, mask, _) in enumerate(tqdm(val_dataloader)):
image = image.to(device)

pred = unet(image)

image = np.squeeze(image.cpu())

gt_labels = np.squeeze(mask.cpu().numpy())

pred = np.squeeze(pred.cpu().detach().numpy())

# feel free to try different thresholds
thresh = threshold_otsu(pred)

# get boundary mask
inner_mask = 0.5 * (pred[0] + pred[1]) > thresh

boundary_distances = distance_transform_edt(inner_mask)

pred_labels = watershed_from_boundary_distance(
boundary_distances, inner_mask, id_offset=0, min_seed_distance=20
)
Expand All @@ -852,7 +870,7 @@ def create_aff_target(self, mask):
# <hr style="height:2px;">
#
# ## Bonus: Further reading on Affinities
# [Here](https://localshapedescriptors.github.io/) is a blog post describing the Local Shape Descriptor method of instance segmentation.
# [Here](https://localshapedescriptors.github.io/) is a blog post describing the Local Shape Descriptor method of instance segmentation.
#
# %% [markdown]
# <hr style="height:2px;">
Expand Down Expand Up @@ -911,4 +929,3 @@ def create_aff_target(self, mask):
print(f"Mean Precision is {np.mean(precision_list):.3f}")
print(f"Mean Recall is {np.mean(recall_list):.3f}")
print(f"Mean Accuracy is {np.mean(accuracy_list):.3f}")

0 comments on commit c706cc8

Please sign in to comment.