def test_show_image_smoke(subtests, mocker, test_image_file, test_image):
mocker.patch("pystiche.image.io._show_pil_image")
image_.show_image(test_image)
with subtests.test(image=test_image_file):
image_.show_image(test_image_file)
with subtests.test(image=None):
with pytest.raises(TypeError):
image_.show_image(None)
with subtests.test(size=100):
image_.show_image(test_image, size=100)
with subtests.test(size=(100, 200)):
image_.show_image(test_image, size=(100, 200))
# # By default the ``edge_sizes`` correspond to the shorter ``edge`` of the images. To # change that you can pass ``edge="long"``. For fine-grained control you can also # pass a sequence comprising ``"short"`` and ``"long"`` to select the ``edge`` for # each level separately. edge_sizes = (300, 550) num_steps = 200 pyramid = ImagePyramid(edge_sizes, num_steps, resize_targets=(criterion, )) print(pyramid) ######################################################################################## # Next up, we load and show the images that will be used in the NST. content_image = images["bird2"].read(device=device) show_image(content_image, title="Input image") ######################################################################################## style_image = images["mosaic"].read(device=device) show_image(style_image, title="Output image") ######################################################################################## # Although the images would be automatically resized during the optimization you might # need to resize them before: if you are working with large source images you might # run out of memory by setting up the targets of the perceptual loss. In that case it # is good practice to resize the images upfront to the largest size the ``pyramid`` # will handle. top_level = pyramid[-1] content_image = top_level.resize_image(content_image)
print(f"I'm working with pystiche=={pystiche.__version__}")
device = get_device()
print(f"I'm working with {device}")
########################################################################################
# In a first step we load and show the images that will be used in the NST.
images = demo_images()
images.download()
size = 500
########################################################################################
content_image = images["castle"].read(size=size, device=device)
show_image(content_image)
########################################################################################
style_image = images["church"].read(size=size, device=device)
show_image(style_image)
########################################################################################
# Unguided image optimization
# ---------------------------
#
# As a baseline we use a default NST with a
# :class:`~pystiche.ops.FeatureReconstructionOperator` as ``content_loss`` and
# :class:`~pystiche.ops.GramOperator` s as ``style_loss``.
multi_layer_encoder = vgg19_multi_layer_encoder()
) criterion = loss.PerceptualLoss(content_loss, style_loss).to(device) print(criterion) ######################################################################################## # Next up, we load and show the images that will be used in the NST. images = demo.images() images.download() size = 500 ######################################################################################## content_image = images["bird2"].read(size=size, device=device) show_image(content_image, title="Content image") criterion.set_content_image(content_image) ######################################################################################## style_image = images["mosaic"].read(size=size, device=device) show_image(style_image, title="Style image") criterion.set_style_image(style_image) ######################################################################################## # Image optimization without pyramid # ---------------------------------- # # As a baseline we use a standard image optimization without pyramid. starting_point = "content"
perceptual_loss = perceptual_loss.to(device) print(perceptual_loss) ######################################################################################## # Training # -------- # # In a first step we load the style image that will be used to train the # ``transformer``. images = demo.images() size = 500 style_image = images["paint"].read(size=size, device=device) show_image(style_image) ######################################################################################## # The training of the ``transformer`` is performed similar to other models in PyTorch. # In every optimization step a batch of content images is drawn from a dataset, which # serve as input for the transformer as well as ``content_image`` for the # ``perceptual_loss``. While the ``style_image`` only has to be set once, the # ``content_image`` has to be reset in every iteration step. # # While this can be done with a boilerplate optimization loop, ``pystiche`` provides # :func:`~pystiche.optim.multi_epoch_model_optimization` that handles the above for you. # # .. note:: # # If the ``perceptual_loss`` is a :class:`~pystiche.loss.PerceptualLoss`, as is the
def test_show_image_size_smoke(patch_show_pil_image, test_image, size):
image_.show_image(test_image, size=size)
def test_show_image_invalid_type():
with pytest.raises(TypeError):
image_.show_image(None)
def test_show_image_file_smoke(patch_show_pil_image, test_image_file):
image_.show_image(test_image_file)
def test_show_image_tensor_smoke(patch_show_pil_image, test_image):
image_.show_image(test_image)
