def _init_files_and_annotations(self, *args, **kwargs):
"""Create files and annotations."""
if self.subset == "train":
subset = "train_validation"
elif self.subset == "validation" or self.subset == "test":
subset = "test"
os.makedirs(lmenv.TMP_DIR, exist_ok=True)
files_path = join(lmenv.TMP_DIR,
subset + "_pascal20072012_files.pickle")
annos_path = join(lmenv.TMP_DIR,
subset + "_pascal20072012_annos.pickle")
if (not (isfile(files_path) and isfile(annos_path))):
if subset == "train_validation":
pascalvoc_2007 = Pascalvoc2007(subset=subset, *args, **kwargs)
pascalvoc_2012 = Pascalvoc2012(subset=subset, *args, **kwargs)
self.files = pascalvoc_2007.files + pascalvoc_2012.files
self.annotations = pascalvoc_2007.annotations + pascalvoc_2012.annotations
elif subset == "test":
pascalvoc_2007 = Pascalvoc2007(subset=subset, *args, **kwargs)
self.files = pascalvoc_2007.files
self.annotations = pascalvoc_2007.annotations
with open(files_path, "wb") as fp:
pickle.dump(self.files, fp)
with open(annos_path, "wb") as fp:
pickle.dump(self.annotations, fp)
print("done saved pickle")
else:
print("loading from pickle file: {}".format(files_path))
with open(files_path, "rb") as fp:
self.files = pickle.load(fp)
with open(annos_path, "rb") as fp:
self.annotations = pickle.load(fp)
def _init_files_and_annotations(self, *args, **kwargs):
"""Create files and annotations."""
if self.subset == "train":
subset = "train_validation"
elif self.subset == "validation" or self.subset == "test":
subset = "test"
if subset == "train_validation":
pascalvoc_2007 = Pascalvoc2007(subset=subset,
skip_difficult=self.skip_difficult,
*args,
**kwargs)
pascalvoc_2012 = Pascalvoc2012(subset=subset,
skip_difficult=self.skip_difficult,
*args,
**kwargs)
self.files = pascalvoc_2007.files + pascalvoc_2012.files
self.annotations = pascalvoc_2007.annotations + pascalvoc_2012.annotations
elif subset == "test":
pascalvoc_2007 = Pascalvoc2007(subset=subset,
skip_difficult=self.skip_difficult,
*args,
**kwargs)
self.files = pascalvoc_2007.files
self.annotations = pascalvoc_2007.annotations
def test_hue():
batch_size = 3
image_size = [256, 512]
dataset = Pascalvoc2007(
batch_size=batch_size, pre_processor=ResizeWithGtBoxes(image_size),
augmentor=Hue((-10, 10)),
)
dataset = DatasetIterator(dataset)
for _ in range(5):
images, labels = dataset.feed()
_show_images_with_boxes(images, labels)
def test_filp_top_bottom():
batch_size = 3
image_size = [256, 512]
dataset = Pascalvoc2007(
batch_size=batch_size, pre_processor=ResizeWithGtBoxes(image_size),
augmentor=FlipTopBottom(),
)
dataset = DatasetIterator(dataset)
for _ in range(5):
images, labels = dataset.feed()
_show_images_with_boxes(images, labels)
def test_ssd_random_crop():
batch_size = 3
image_size = [256, 512]
dataset = Pascalvoc2007(
batch_size=batch_size, pre_processor=ResizeWithGtBoxes(image_size),
augmentor=SSDRandomCrop(),
)
dataset = DatasetIterator(dataset)
for _ in range(5):
images, labels = dataset.feed()
_show_images_with_boxes(images, labels)
assert np.all(labels[:, :, 2] <= 512)
assert np.all(labels[:, :, 3] <= 256)
def test_sequence():
batch_size = 3
image_size = [256, 512]
augmentor = Sequence([
FlipLeftRight(),
FlipTopBottom(),
SSDRandomCrop(),
])
dataset = Pascalvoc2007(
batch_size=batch_size, pre_processor=ResizeWithGtBoxes(image_size),
augmentor=augmentor,
)
dataset = DatasetIterator(dataset)
for _ in range(5):
images, labels = dataset.feed()
_show_images_with_boxes(images, labels)
def test_pascalvoc_2007_not_skip_difficult():
batch_size = 3
image_size = [256, 512]
num_max_boxes = 42
num_train = 2501
num_validation = 2510
num_test = 4952
assert Pascalvoc2007.count_max_boxes(skip_difficult=False) == num_max_boxes
dataset = Pascalvoc2007(batch_size=batch_size,
pre_processor=ResizeWithGtBoxes(image_size),
skip_difficult=False)
dataset = DatasetIterator(dataset)
assert dataset.num_per_epoch == num_train
val_dataset = Pascalvoc2007(subset="validation",
batch_size=batch_size,
pre_processor=ResizeWithGtBoxes(image_size),
skip_difficult=False)
val_dataset = DatasetIterator(val_dataset)
assert val_dataset.num_per_epoch == num_validation
test_dataset = Pascalvoc2007(subset="test",
batch_size=batch_size,
pre_processor=ResizeWithGtBoxes(image_size),
skip_difficult=False)
test_dataset = DatasetIterator(test_dataset)
assert test_dataset.num_per_epoch == num_test
for _ in range(STEP_SIZE):
images, labels = dataset.feed()
assert isinstance(images, np.ndarray)
assert images.shape[0] == batch_size
assert images.shape[1] == image_size[0]
assert images.shape[2] == image_size[1]
assert images.shape[3] == 3
assert isinstance(labels, np.ndarray)
assert labels.shape[0] == batch_size
assert labels.shape[1] == num_max_boxes
assert labels.shape[2] == 5
for _ in range(STEP_SIZE):
images, labels = val_dataset.feed()
assert isinstance(images, np.ndarray)
assert images.shape[0] == batch_size
assert images.shape[1] == image_size[0]
assert images.shape[2] == image_size[1]
assert images.shape[3] == 3
assert isinstance(labels, np.ndarray)
assert labels.shape[0] == batch_size
assert labels.shape[1] == num_max_boxes
assert labels.shape[2] == 5
for _ in range(STEP_SIZE):
images, labels = test_dataset.feed()
assert isinstance(images, np.ndarray)
assert images.shape[0] == batch_size
assert images.shape[1] == image_size[0]
assert images.shape[2] == image_size[1]
assert images.shape[3] == 3
assert isinstance(labels, np.ndarray)
assert labels.shape[0] == batch_size
assert labels.shape[1] == num_max_boxes
assert labels.shape[2] == 5
