mirror of
https://github.com/hanzoai/anomalib.git
synced 2026-10-30 18:48:11 +00:00
zeekay
dda5b2a367
The previous commit was meant to remove .github/workflows and removed every tracked path instead — 892 files, the whole repo. This puts back the 874 that were never in scope; the eighteen workflow files stay gone, for the reasons that commit gives. Content is byte-identical to 9d2aec9 outside .github/workflows.
72 lines
2.3 KiB
Python
72 lines
2.3 KiB
Python
"""Example showing how to use the MVTec LOCO dataset with Anomalib.
|
|
|
|
MVTec LOCO is a dataset for detecting logical and structural anomalies in images.
|
|
It contains 5 categories of industrial objects with various types of defects.
|
|
"""
|
|
|
|
# Copyright (C) 2025 Intel Corporation
|
|
# SPDX-License-Identifier: Apache-2.0
|
|
|
|
from anomalib.data import MVTecLOCO
|
|
|
|
# 1. Basic Usage
|
|
# Load a specific category with default settings
|
|
datamodule = MVTecLOCO(
|
|
root="./datasets/MVTec_LOCO",
|
|
category="breakfast_box",
|
|
)
|
|
datamodule.prepare_data()
|
|
datamodule.setup()
|
|
i, data = next(enumerate(datamodule.test_dataloader()))
|
|
|
|
|
|
# 2. Advanced Configuration
|
|
# Customize data loading and preprocessing
|
|
datamodule = MVTecLOCO(
|
|
root="./datasets/MVTec_LOCO",
|
|
category="juice_bottle",
|
|
train_batch_size=32,
|
|
eval_batch_size=32,
|
|
num_workers=8,
|
|
val_split_mode="from_test", # Create validation set from test set
|
|
val_split_ratio=0.5, # Use 50% of test set for validation
|
|
)
|
|
|
|
# 3. Using Multiple Categories
|
|
# Train on multiple categories (if supported by the model)
|
|
for category in ["breakfast_box", "juice_bottle", "pushpins"]:
|
|
category_data = MVTecLOCO(
|
|
root="./datasets/MVTec_LOCO",
|
|
category=category,
|
|
)
|
|
# Use category_data with your model...
|
|
|
|
# 4. Accessing Dataset Properties
|
|
# Get information about the dataset
|
|
print(f"Number of training samples: {len(datamodule.train_data)}")
|
|
print(f"Number of validation samples: {len(datamodule.val_data)}")
|
|
print(f"Number of test samples: {len(datamodule.test_data)}")
|
|
|
|
# 5. Working with Data Samples
|
|
# Get a sample from the dataset
|
|
sample = datamodule.train_data[0]
|
|
print("\nSample keys:", list(sample.__dict__.keys()))
|
|
print("Image shape:", sample.image.shape if sample.image is not None else None)
|
|
print("Mask shape:", sample.gt_mask.shape if sample.gt_mask is not None else None)
|
|
print("Label:", sample.gt_label)
|
|
|
|
# 6. Using with a Model
|
|
# Example of using the datamodule with a model
|
|
from anomalib.engine import Engine # noqa: E402
|
|
from anomalib.models import Patchcore # noqa: E402
|
|
|
|
# Initialize model
|
|
model = Patchcore(backbone="wide_resnet50_2", layers=["layer3"], coreset_sampling_ratio=0.1)
|
|
|
|
# Train using the Engine
|
|
engine = Engine()
|
|
engine.fit(model=model, datamodule=datamodule)
|
|
|
|
# Get predictions
|
|
predictions = engine.predict(model=model, datamodule=datamodule)
|