def measure_model(
model_version,
dataset,
out_folder,
weights_dir,
device,
method=METHODS["gradcam"],
sample_images=50,
step=1,
):
invTrans = get_inverse_normalization_transformation()
data_dir = os.path.join("data")
if model_version == "resnet18":
model = create_resnet18_model(num_of_classes=NUM_OF_CLASSES[dataset])
elif model_version == "resnet50":
model = create_resnet50_model(num_of_classes=NUM_OF_CLASSES[dataset])
elif model_version == "densenet":
model = create_densenet121_model(
num_of_classes=NUM_OF_CLASSES[dataset])
else:
model = create_efficientnetb0_model(
num_of_classes=NUM_OF_CLASSES[dataset])
model.load_state_dict(torch.load(weights_dir))
# print(model)
model.eval()
model.to(device)
test_dataset = CustomDataset(
dataset=dataset,
transformer=get_default_transformation(),
data_type="test",
root_dir=data_dir,
step=step,
)
data_loader = torch.utils.data.DataLoader(test_dataset,
batch_size=1,
shuffle=False,
num_workers=4)
try:
image_ids = random.sample(range(0, test_dataset.__len__()),
sample_images)
except ValueError:
raise ValueError(
f"Image sample number ({sample_images}) exceeded dataset size ({test_dataset.__len__()})."
)
classes_map = test_dataset.classes_map
print(f"Measuring {model_version} on {dataset} dataset, with {method}")
print("-" * 10)
pbar = tqdm(total=test_dataset.__len__(), desc="Model test completion")
multipy_by_inputs = False
if method == METHODS["ig"]:
attr_method = IntegratedGradients(model)
nt_samples = 8
n_perturb_samples = 3
if method == METHODS["saliency"]:
attr_method = Saliency(model)
nt_samples = 8
n_perturb_samples = 10
if method == METHODS["gradcam"]:
if model_version == "efficientnet":
attr_method = GuidedGradCam(model, model._conv_stem)
elif model_version == "densenet":
attr_method = GuidedGradCam(model, model.features.conv0)
else:
attr_method = GuidedGradCam(model, model.conv1)
nt_samples = 8
n_perturb_samples = 10
if method == METHODS["deconv"]:
attr_method = Deconvolution(model)
nt_samples = 8
n_perturb_samples = 10
if method == METHODS["gradshap"]:
attr_method = GradientShap(model)
nt_samples = 8
if model_version == "efficientnet":
n_perturb_samples = 3
elif model_version == "densenet":
n_perturb_samples = 2
else:
n_perturb_samples = 10
if method == METHODS["gbp"]:
attr_method = GuidedBackprop(model)
nt_samples = 8
n_perturb_samples = 10
if method == "lime":
attr_method = Lime(model)
nt_samples = 8
n_perturb_samples = 10
feature_mask = torch.tensor(lime_mask).to(device)
multipy_by_inputs = True
if method == METHODS['ig']:
nt = attr_method
else:
nt = NoiseTunnel(attr_method)
scores = []
@infidelity_perturb_func_decorator(multipy_by_inputs=multipy_by_inputs)
def perturb_fn(inputs):
noise = torch.tensor(np.random.normal(0, 0.003, inputs.shape)).float()
noise = noise.to(device)
return inputs - noise
for input, label in data_loader:
pbar.update(1)
inv_input = invTrans(input)
input = input.to(device)
input.requires_grad = True
output = model(input)
output = F.softmax(output, dim=1)
prediction_score, pred_label_idx = torch.topk(output, 1)
prediction_score = prediction_score.cpu().detach().numpy()[0][0]
pred_label_idx.squeeze_()
if method == METHODS['gradshap']:
baseline = torch.randn(input.shape)
baseline = baseline.to(device)
if method == "lime":
attributions = attr_method.attribute(input, target=1, n_samples=50)
elif method == METHODS['ig']:
attributions = nt.attribute(
input,
target=pred_label_idx,
n_steps=25,
)
elif method == METHODS['gradshap']:
attributions = nt.attribute(input,
target=pred_label_idx,
baselines=baseline)
else:
attributions = nt.attribute(
input,
nt_type="smoothgrad",
nt_samples=nt_samples,
target=pred_label_idx,
)
infid = infidelity(model,
perturb_fn,
input,
attributions,
target=pred_label_idx)
if method == "lime":
sens = sensitivity_max(
attr_method.attribute,
input,
target=pred_label_idx,
n_perturb_samples=1,
n_samples=200,
feature_mask=feature_mask,
)
elif method == METHODS['ig']:
sens = sensitivity_max(
nt.attribute,
input,
target=pred_label_idx,
n_perturb_samples=n_perturb_samples,
n_steps=25,
)
elif method == METHODS['gradshap']:
sens = sensitivity_max(nt.attribute,
input,
target=pred_label_idx,
n_perturb_samples=n_perturb_samples,
baselines=baseline)
else:
sens = sensitivity_max(
nt.attribute,
input,
target=pred_label_idx,
n_perturb_samples=n_perturb_samples,
)
inf_value = infid.cpu().detach().numpy()[0]
sens_value = sens.cpu().detach().numpy()[0]
if pbar.n in image_ids:
attr_data = attributions.squeeze().cpu().detach().numpy()
fig, ax = viz.visualize_image_attr_multiple(
np.transpose(attr_data, (1, 2, 0)),
np.transpose(inv_input.squeeze().cpu().detach().numpy(),
(1, 2, 0)),
["original_image", "heat_map"],
["all", "positive"],
titles=["original_image", "heat_map"],
cmap=default_cmap,
show_colorbar=True,
use_pyplot=False,
fig_size=(8, 6),
)
ax[0].set_xlabel(
f"Infidelity: {'{0:.6f}'.format(inf_value)}\n Sensitivity: {'{0:.6f}'.format(sens_value)}"
)
fig.suptitle(
f"True: {classes_map[str(label.numpy()[0])][0]}, Pred: {classes_map[str(pred_label_idx.item())][0]}\nScore: {'{0:.4f}'.format(prediction_score)}",
fontsize=16,
)
fig.savefig(
os.path.join(
out_folder,
f"{str(pbar.n)}-{classes_map[str(label.numpy()[0])][0]}-{classes_map[str(pred_label_idx.item())][0]}.png",
))
plt.close(fig)
# if pbar.n > 25:
# break
scores.append([inf_value, sens_value])
pbar.close()
np.savetxt(
os.path.join(out_folder, f"{model_version}-{dataset}-{method}.csv"),
np.array(scores),
delimiter=",",
header="infidelity,sensitivity",
)
print(f"Artifacts stored at {out_folder}")
def train_resnet(dataset, model_type="resnet18", data_dir=None, skip=None):
data_transform = get_default_transformation()
train_dataset = CustomDataset(
dataset=dataset,
transformer=data_transform,
data_type="train",
root_dir=data_dir,
skip=skip,
)
test_dataset = CustomDataset(dataset=dataset,
transformer=data_transform,
data_type="test",
root_dir=data_dir)
# def show_image(image, label, dataset):
# print(f"Label: {label}")
# plt.imshow(image.permute(1,2,0))
# plt.show()
#
#
# show_image(*train_dataset[0], train_dataset)
dataset_loaders = {
"train":
torch.utils.data.DataLoader(train_dataset,
batch_size=8,
shuffle=True,
num_workers=4),
"val":
torch.utils.data.DataLoader(test_dataset,
batch_size=8,
shuffle=False,
num_workers=4),
}
print(f"Num of classes {len(train_dataset.classes)}")
if model_type == "resnet18":
model_ft = create_resnet18_model(
num_of_classes=len(train_dataset.classes))
else:
model_ft = create_resnet50_model(
num_of_classes=len(train_dataset.classes))
model_ft = model_ft.to(device)
criterion = nn.CrossEntropyLoss()
# Observe that all parameters are being optimized
optimizer_ft = optim.SGD(model_ft.parameters(), lr=0.001, momentum=0.9)
# Decay LR by a factor of 0.1 every 7 epochs
exp_lr_scheduler = lr_scheduler.StepLR(optimizer_ft,
step_size=7,
gamma=0.1)
model_ft = train_model(
model_ft,
dataset_loaders,
criterion,
optimizer_ft,
exp_lr_scheduler,
device,
num_epochs=15,
)
return model_ft
def measure_filter_model(
model_version,
dataset,
out_folder,
weights_dir,
device,
method=METHODS["gradcam"],
sample_images=50,
step=1,
use_infidelity=False,
use_sensitivity=False,
render=False,
ids=None,
):
invTrans = get_inverse_normalization_transformation()
data_dir = os.path.join("data")
if model_version == "resnet18":
model = create_resnet18_model(num_of_classes=NUM_OF_CLASSES[dataset])
elif model_version == "resnet50":
model = create_resnet50_model(num_of_classes=NUM_OF_CLASSES[dataset])
elif model_version == "densenet":
model = create_densenet121_model(
num_of_classes=NUM_OF_CLASSES[dataset])
else:
model = create_efficientnetb0_model(
num_of_classes=NUM_OF_CLASSES[dataset])
model.load_state_dict(torch.load(weights_dir))
# print(model)
model.eval()
model.to(device)
test_dataset = CustomDataset(
dataset=dataset,
transformer=get_default_transformation(),
data_type="test",
root_dir=data_dir,
step=step,
add_filters=True,
ids=ids,
)
data_loader = torch.utils.data.DataLoader(test_dataset,
batch_size=1,
shuffle=False,
num_workers=4)
try:
image_ids = random.sample(range(0, test_dataset.__len__()),
test_dataset.__len__())
except ValueError:
raise ValueError(
f"Image sample number ({test_dataset.__len__()}) exceeded dataset size ({test_dataset.__len__()})."
)
classes_map = test_dataset.classes_map
print(f"Measuring {model_version} on {dataset} dataset, with {method}")
print("-" * 10)
pbar = tqdm(total=test_dataset.__len__(), desc="Model test completion")
multipy_by_inputs = False
if method == METHODS["ig"]:
attr_method = IntegratedGradients(model)
nt_samples = 1
n_perturb_samples = 1
if method == METHODS["saliency"]:
attr_method = Saliency(model)
nt_samples = 8
n_perturb_samples = 2
if method == METHODS["gradcam"]:
if model_version == "efficientnet":
attr_method = GuidedGradCam(model, model._conv_stem)
elif model_version == "densenet":
attr_method = GuidedGradCam(model, model.features.conv0)
else:
attr_method = GuidedGradCam(model, model.conv1)
nt_samples = 8
n_perturb_samples = 2
if method == METHODS["deconv"]:
attr_method = Deconvolution(model)
nt_samples = 8
n_perturb_samples = 2
if method == METHODS["gradshap"]:
attr_method = GradientShap(model)
nt_samples = 8
n_perturb_samples = 2
if method == METHODS["gbp"]:
attr_method = GuidedBackprop(model)
nt_samples = 8
n_perturb_samples = 2
if method == "lime":
attr_method = Lime(model)
nt_samples = 8
n_perturb_samples = 2
feature_mask = torch.tensor(lime_mask).to(device)
multipy_by_inputs = True
if method == METHODS["ig"]:
nt = attr_method
else:
nt = NoiseTunnel(attr_method)
scores = []
@infidelity_perturb_func_decorator(multipy_by_inputs=multipy_by_inputs)
def perturb_fn(inputs):
noise = torch.tensor(np.random.normal(0, 0.003, inputs.shape)).float()
noise = noise.to(device)
return inputs - noise
OUR_FILTERS = [
"none",
"fx_freaky_details 2,10,1,11,0,32,0",
"normalize_local 8,10",
"fx_boost_chroma 90,0,0",
"fx_mighty_details 25,1,25,1,11,0",
"sharpen 300",
]
idx = 0
filter_count = 0
filter_attrs = {filter_name: [] for filter_name in OUR_FILTERS}
predicted_main_class = 0
for input, label in data_loader:
pbar.update(1)
inv_input = invTrans(input)
input = input.to(device)
input.requires_grad = True
output = model(input)
output = F.softmax(output, dim=1)
prediction_score, pred_label_idx = torch.topk(output, 1)
prediction_score = prediction_score.cpu().detach().numpy()[0][0]
pred_label_idx.squeeze_()
if OUR_FILTERS[filter_count] == 'none':
predicted_main_class = pred_label_idx.item()
if method == METHODS["gradshap"]:
baseline = torch.randn(input.shape)
baseline = baseline.to(device)
if method == "lime":
attributions = attr_method.attribute(input, target=1, n_samples=50)
elif method == METHODS["ig"]:
attributions = nt.attribute(
input,
target=predicted_main_class,
n_steps=25,
)
elif method == METHODS["gradshap"]:
attributions = nt.attribute(input,
target=predicted_main_class,
baselines=baseline)
else:
attributions = nt.attribute(
input,
nt_type="smoothgrad",
nt_samples=nt_samples,
target=predicted_main_class,
)
if use_infidelity:
infid = infidelity(model,
perturb_fn,
input,
attributions,
target=predicted_main_class)
inf_value = infid.cpu().detach().numpy()[0]
else:
inf_value = 0
if use_sensitivity:
if method == "lime":
sens = sensitivity_max(
attr_method.attribute,
input,
target=predicted_main_class,
n_perturb_samples=1,
n_samples=200,
feature_mask=feature_mask,
)
elif method == METHODS["ig"]:
sens = sensitivity_max(
nt.attribute,
input,
target=predicted_main_class,
n_perturb_samples=n_perturb_samples,
n_steps=25,
)
elif method == METHODS["gradshap"]:
sens = sensitivity_max(
nt.attribute,
input,
target=predicted_main_class,
n_perturb_samples=n_perturb_samples,
baselines=baseline,
)
else:
sens = sensitivity_max(
nt.attribute,
input,
target=predicted_main_class,
n_perturb_samples=n_perturb_samples,
)
sens_value = sens.cpu().detach().numpy()[0]
else:
sens_value = 0
# filter_name = test_dataset.data.iloc[pbar.n]["filter"].split(" ")[0]
attr_data = attributions.squeeze().cpu().detach().numpy()
if render:
fig, ax = viz.visualize_image_attr_multiple(
np.transpose(attr_data, (1, 2, 0)),
np.transpose(inv_input.squeeze().cpu().detach().numpy(),
(1, 2, 0)),
["original_image", "heat_map"],
["all", "positive"],
titles=["original_image", "heat_map"],
cmap=default_cmap,
show_colorbar=True,
use_pyplot=False,
fig_size=(8, 6),
)
if use_sensitivity or use_infidelity:
ax[0].set_xlabel(
f"Infidelity: {'{0:.6f}'.format(inf_value)}\n Sensitivity: {'{0:.6f}'.format(sens_value)}"
)
fig.suptitle(
f"True: {classes_map[str(label.numpy()[0])][0]}, Pred: {classes_map[str(pred_label_idx.item())][0]}\nScore: {'{0:.4f}'.format(prediction_score)}",
fontsize=16,
)
fig.savefig(
os.path.join(
out_folder,
f"{str(idx)}-{str(filter_count)}-{str(label.numpy()[0])}-{str(OUR_FILTERS[filter_count])}-{classes_map[str(label.numpy()[0])][0]}-{classes_map[str(pred_label_idx.item())][0]}.png",
))
plt.close(fig)
# if pbar.n > 25:
# break
score_for_true_label = output.cpu().detach().numpy(
)[0][predicted_main_class]
filter_attrs[OUR_FILTERS[filter_count]] = [
np.moveaxis(attr_data, 0, -1),
"{0:.8f}".format(score_for_true_label),
]
data_range_for_current_set = MAX_ATT_VALUES[model_version][method][
dataset]
filter_count += 1
if filter_count >= len(OUR_FILTERS):
ssims = []
for rot in OUR_FILTERS:
ssims.append("{0:.8f}".format(
ssim(
filter_attrs["none"][0],
filter_attrs[rot][0],
win_size=11,
data_range=data_range_for_current_set,
multichannel=True,
)))
ssims.append(filter_attrs[rot][1])
scores.append(ssims)
filter_count = 0
predicted_main_class = 0
idx += 1
pbar.close()
indexes = []
for filter_name in OUR_FILTERS:
indexes.append(str(filter_name) + "-ssim")
indexes.append(str(filter_name) + "-score")
np.savetxt(
os.path.join(
out_folder,
f"{model_version}-{dataset}-{method}-ssim-with-range.csv"),
np.array(scores),
delimiter=";",
fmt="%s",
header=";".join([str(rot) for rot in indexes]),
)
print(f"Artifacts stored at {out_folder}")
def test_model(model_version,
dataset,
out_folder,
weights_dir,
device,
version="100%"):
data_dir = os.path.join("data")
if model_version == "resnet18":
model = create_resnet18_model(num_of_classes=NUM_OF_CLASSES[dataset])
elif model_version == "resnet50":
model = create_resnet50_model(num_of_classes=NUM_OF_CLASSES[dataset])
elif model_version == "densenet":
model = create_densenet121_model(
num_of_classes=NUM_OF_CLASSES[dataset])
else:
model = create_efficientnetb0_model(
num_of_classes=NUM_OF_CLASSES[dataset])
model.load_state_dict(torch.load(weights_dir))
model.eval()
model.to(device)
test_dataset = CustomDataset(
dataset=dataset,
transformer=get_default_transformation(),
data_type="test",
root_dir=data_dir,
)
data_loader = torch.utils.data.DataLoader(test_dataset,
batch_size=1,
shuffle=False,
num_workers=4)
preds = []
labels = []
scores = []
print("-" * 30)
pbar = tqdm(total=test_dataset.__len__(), desc="Model test completion")
for input, label in data_loader:
input = input.to(device)
pbar.update(1)
output = model(input)
output = F.softmax(output, dim=1)
prediction_score, pred_label_idx = torch.topk(output, 1)
pred_label_idx.squeeze_()
label_val = int(label.detach().numpy()[0])
pred_val = int(pred_label_idx.detach().item())
labels.append(label_val)
preds.append(pred_val)
scores.append([
label_val,
pred_val,
label_val == pred_val,
prediction_score.detach().cpu().numpy()[0][0],
])
pbar.close()
conf_matrix, class_report = run_eval(preds, labels, test_dataset.classes)
f1 = "{0:.4f}".format(class_report["weighted avg"]["f1-score"])
acc = "{0:.4f}".format(class_report["weighted avg"]["precision"])
subtitle = f"F1: {f1}, Prec: {acc}"
save_cm(
conf_matrix,
test_dataset.classes,
os.path.join(out_folder, f"{model_version}-{dataset}-{version}.png"),
subtitle=subtitle,
)
classification_report_latex(
class_report,
filename=os.path.join(out_folder,
f"{model_version}-{dataset}-{version}.txt"),
)
with open(
os.path.join(out_folder,
f"{model_version}-{dataset}-{version}.csv"),
"w") as tf:
tf.write(f"{f1},{acc}")
scores_df = pd.DataFrame(np.array(scores),
columns=['true', 'pred', 'currect', 'score'])
scores_df.to_csv(
os.path.join(out_folder,
f"{model_version}-{dataset}-{version}-scores.csv"))
print(
f'Artifacts stored at {os.path.join(out_folder, f"{model_version}-{dataset}-{version}")}.*'
)