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.
93 lines
3.0 KiB
Python
93 lines
3.0 KiB
Python
"""Test PRO metric."""
|
|
|
|
# Copyright (C) 2023-2024 Intel Corporation
|
|
# SPDX-License-Identifier: Apache-2.0
|
|
|
|
import pytest
|
|
import torch
|
|
from torchvision.transforms import RandomAffine
|
|
|
|
from anomalib.data.utils.generators.perlin import generate_perlin_noise
|
|
from anomalib.metrics.pro import _PRO as PRO
|
|
from anomalib.metrics.pro import connected_components_cpu, connected_components_gpu
|
|
|
|
|
|
def test_pro() -> None:
|
|
"""Checks if PRO metric computes the (macro) average of the per-region overlap."""
|
|
labels = torch.Tensor(
|
|
[
|
|
[
|
|
[0, 0, 0, 0, 0],
|
|
[1, 0, 0, 0, 0],
|
|
[0, 0, 0, 0, 0],
|
|
[1, 1, 0, 0, 0],
|
|
[0, 0, 0, 0, 0],
|
|
[1, 1, 1, 0, 0],
|
|
[0, 0, 0, 0, 0],
|
|
[1, 1, 1, 1, 0],
|
|
[0, 0, 0, 0, 0],
|
|
[1, 1, 1, 1, 1],
|
|
],
|
|
],
|
|
)
|
|
# ground truth mask is int type
|
|
labels = labels.type(torch.int32)
|
|
|
|
preds = (torch.arange(10) / 10) + 0.05
|
|
# metrics receive squeezed predictions (N, H, W)
|
|
preds = preds.unsqueeze(1).repeat(1, 5).view(1, 10, 5)
|
|
|
|
thresholds = [0.0, 0.2, 0.4, 0.6, 0.8, 1.0]
|
|
targets = [1.0, 0.8, 0.6, 0.4, 0.2, 0.0]
|
|
for threshold, target in zip(thresholds, targets, strict=True):
|
|
pro = PRO(threshold=threshold)
|
|
pro.update(preds, labels)
|
|
assert pro.compute() == target
|
|
|
|
|
|
@pytest.mark.gpu
|
|
def test_device_consistency() -> None:
|
|
"""Test if the pro metric yields the same results between cpu and gpu.
|
|
|
|
Note: This test will only run on a GPU-enabled device.
|
|
"""
|
|
transform = RandomAffine(5, None, (0.95, 1.05), 5)
|
|
|
|
batch = torch.zeros((32, 256, 256))
|
|
for i in range(batch.shape[0]):
|
|
batch[i, ...] = generate_perlin_noise(256, 256, scale=(4, 4)) > 0.5
|
|
# ground truth mask is int type
|
|
batch = batch.type(torch.int32)
|
|
|
|
preds = transform(batch)
|
|
|
|
pro_cpu = PRO()
|
|
pro_gpu = PRO()
|
|
|
|
pro_cpu.update(preds.cpu(), batch.cpu())
|
|
pro_gpu.update(preds.cuda(), batch.cuda())
|
|
|
|
assert torch.isclose(pro_cpu.compute(), pro_gpu.compute().cpu())
|
|
|
|
|
|
@pytest.mark.gpu
|
|
def test_connected_component_labeling() -> None:
|
|
"""Tests if the connected component labeling algorithms on cpu and gpu yield the same result.
|
|
|
|
Note: This test will only run on a GPU-enabled device.
|
|
"""
|
|
# generate batch of random binary images using perlin noise
|
|
batch = torch.zeros((32, 1, 256, 256))
|
|
for i in range(batch.shape[0]):
|
|
batch[i, ...] = generate_perlin_noise(256, 256, scale=(4, 4)) > 0.5
|
|
|
|
# get connected component results on both cpu and gpu
|
|
cc_cpu = connected_components_cpu(batch.cpu())
|
|
cc_gpu = connected_components_gpu(batch.cuda())
|
|
|
|
# check if comps are ordered from 0 to N
|
|
assert len(cc_cpu.unique()) == cc_cpu.unique().max() + 1
|
|
assert len(cc_gpu.unique()) == cc_gpu.unique().max() + 1
|
|
# check if same number of comps found between cpu and gpu
|
|
assert len(cc_cpu.unique()) == len(cc_gpu.unique())
|