Skip to content

ENH: Adds BioViL-T Linear Classifier (Chexpert Classes) #860

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.

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

Draft
wants to merge 20 commits into
base: main
Choose a base branch
from
Draft
Show file tree
Hide file tree
Changes from all commits
Commits
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
35 changes: 35 additions & 0 deletions hi-ml-multimodal/src/health_multimodal/image/model/pretrained.py
Original file line number Diff line number Diff line change
Expand Up @@ -83,4 +83,39 @@ def get_biovil_t_image_encoder() -> ImageModel:
joint_feature_size=JOINT_FEATURE_SIZE,
pretrained_model_path=biovilt_checkpoint_path,
)


def get_biovil_t_linear_image_classifier(biovilt_checkpoint_path: str) -> ImageModel:
"""
Download weights from Hugging Face and instantiate the image model.

The model is initialized with a linear classifier on top of the
BiomedVLP-BioViL-T image encoder.

Binary classification tasks in order:
['Enlarged Cardiomediastinum', 'Cardiomegaly', 'Edema', 'Consolidation',
'Pneumonia', 'Pneumothorax', 'Pleural Effusion', 'No Finding']

:param biovilt_checkpoint_path: Path to the checkpoint file.

Example:
>>> checkpoint_path = "..."
>>> image_model = get_biovil_t_linear_image_classifier(checkpoint_path)
>>> image_model(torch.Tensor(batch_size, 3, 448, 448)).class_logits.shape
torch.Size([batch_size, num_classes, num_tasks])
"""

num_classes = 2
num_tasks = 8

model_type = ImageEncoderType.RESNET50_MULTI_IMAGE
image_model = ImageModel(
img_encoder_type=model_type,
joint_feature_size=JOINT_FEATURE_SIZE,
pretrained_model_path=biovilt_checkpoint_path,
num_classes=num_classes,
num_tasks=num_tasks,
classifier_hidden_dim=None,
)

return image_model
42 changes: 42 additions & 0 deletions hi-ml-multimodal/test_multimodal/image/model/test_pretrained.py
Original file line number Diff line number Diff line change
@@ -0,0 +1,42 @@
import pytest
import torch

from health_multimodal.image.model.model import ImageModel
from health_multimodal.image.model.pretrained import get_biovil_t_linear_image_classifier


@pytest.fixture
def dummy_input_tensor() -> torch.Tensor:
batch_size = 2
return torch.randn(batch_size, 3, 448, 448)


@pytest.fixture
def biovil_t_linear_image_classifier() -> ImageModel:
# Set the path to None to initialize the model weights with random values
biovil_t_checkpoint_path = None
return get_biovil_t_linear_image_classifier(biovil_t_checkpoint_path)


def test_get_biovil_t_linear_image_classifier_shape(
biovil_t_linear_image_classifier: ImageModel, dummy_input_tensor: torch.Tensor
) -> None:
num_classes = 2
num_tasks = 8

output = biovil_t_linear_image_classifier(dummy_input_tensor)
expected_shape = torch.Size([dummy_input_tensor.shape[0], num_classes, num_tasks])

assert (
output.class_logits.shape == expected_shape
), f"Unexpected output shape. Expected {expected_shape} but got {output.class_logits.shape}"


def test_get_biovil_t_linear_image_classifier_inference(
biovil_t_linear_image_classifier: ImageModel, dummy_input_tensor: torch.Tensor
) -> None:
with torch.no_grad():
try:
_ = biovil_t_linear_image_classifier(dummy_input_tensor)
except Exception as e:
pytest.fail(f"Model inference failed: {e}")