Skip to content

feat: gnn-based inference #125

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
anna-grim merged 1 commit into from
Apr 16, 2024
Merged

feat: gnn-based inference #125

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
1 change: 0 additions & 1 deletion src/deep_neurographs/machine_learning/feature_generation.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -335,7 +335,6 @@ def generate_skel_features(neurograph, proposals):
n_nearby_leafs(neurograph, proposal),
get_radii(neurograph, proposal),
get_avg_radii(neurograph, proposal),
get_avg_branch_lens(neurograph, proposal),
get_directionals(neurograph, proposal, 8),
get_directionals(neurograph, proposal, 16),
get_directionals(neurograph, proposal, 32),
Expand Down
12 changes: 10 additions & 2 deletions src/deep_neurographs/machine_learning/inference.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -156,7 +156,7 @@ def predict(
confidence_threshold=0.7,
):
# Generate features
features = feature_generation.run_on_proposals(
features = feature_generation.run(
neurograph,
model_type,
search_radius,
Expand All @@ -167,18 +167,26 @@ def predict(
dataset = ml_utils.init_dataset(neurograph, features, model_type)

# Run model
idx_to_edge = get_idxs(dataset, model_type)
proposal_probs = run_model(dataset, model, model_type)
accepts, graph = build.get_accepted_proposals(
neurograph,
graph,
proposal_probs,
dataset["idx_to_edge"],
idx_to_edge,
high_threshold=0.95,
low_threshold=confidence_threshold,
)
return accepts, graph


def get_idxs(dataset, model_type):
if "Graph" in model_type:
return dataset.idxs_proposals["idx_to_edge"]
else:
return dataset["idx_to_edge"]


# -- Whole Brain Seed-Based Inference --
def build_from_soma(
neurograph, labels_path, chunk_origin, chunk_shape=CHUNK_SHAPE, n_hops=1
Expand Down
18 changes: 17 additions & 1 deletion src/deep_neurographs/machine_learning/ml_utils.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -15,7 +15,7 @@
import torch
from sklearn.ensemble import AdaBoostClassifier, RandomForestClassifier

from deep_neurographs.machine_learning import feature_generation
from deep_neurographs.machine_learning import feature_generation, graph_datasets
from deep_neurographs.machine_learning.datasets import (
ImgProposalDataset,
MultiModalDataset,
Expand Down Expand Up @@ -152,6 +152,22 @@ def get_dataset(inputs, targets, model_type, transform, lengths):

def init_dataset(
neurographs, features, model_type, block_ids=None, transform=False
):
if "Graph" in model_type:
dataset = graph_datasets.init(neurographs, features)
else:
dataset = init_proposal_dataset(
neurographs,
features,
model_type,
block_ids=block_ids,
transform=transform
)
return dataset


def init_proposal_dataset(
neurographs, features, model_type, block_ids=None, transform=False
):
# Extract features
inputs, targets, idx_transforms = feature_generation.get_matrix(
Expand Down