I am using albumentations for my object detection data augmentation because it deals with bbox at the same time. My format is [xmin, ymin, xmax, ymax] with pascal_voc format.
A.Compose(
[
A.HorizontalFlip(p=0.5),
A.VerticalFlip(p=0.5),
A.OneOf(
[
A.ShiftScaleRotate(
shift_limit=0.2,
scale_limit=(-0.1, 0.6),
rotate_limit=0,
p=0.5,
),
A.RandomSizedCrop(
min_max_height=[200, 352], size=(352, 2048), p=0.5
),
],
p=0.5,
),
A.AdditiveNoise(
noise_type="uniform", noise_params={"ranges":[(-0.2, 0.2)]}, p=0.5
),
ToTensorV2(),
],
bbox_params=A.BboxParams(
format="pascal_voc", label_fields=["class_labels"]
),
)
Then I create my data as:
def __getitem__(self, index):
data = get_data(index)
torch_target = _to_torch_target(data)
torch_img = _get_torch_image(data)
if self.transform:
sample = self.transform(
image=torch_img.numpy(),
bboxes=torch_target["boxes"],
class_labels=torch_target["labels"],
)
torch_img = sample["image"]
torch_target["boxes"] = torch.Tensor(sample["bboxes"])
torch_target["labels"] = torch.Tensor(sample["class_labels"])
torch_target["boxes"] = torch.as_tensor(torch_target["boxes"], dtype=torch.float32)
torch_target["labels"] = torch.as_tensor(torch_target["labels"], dtype=torch.int64)
return torch_img, torch_target
But I am getting an error:
Traceback (most recent call last):
File "/opt/conda/lib/python3.10/runpy.py", line 196, in _run_module_as_main
return _run_code(code, main_globals, None,
File "/opt/conda/lib/python3.10/runpy.py", line 86, in _run_code
exec(code, run_globals)
File "/home/.../main.py", line 116, in <module>
train(results_path, dataset_path, weights_path, **conf)
File "/home/.../main.py", line 77, in train
trainer.fit(model, train_dataloader, val_dataloader)
File "/opt/conda/lib/python3.10/site-packages/lightning/pytorch/trainer/trainer.py", line 543, in fit
call._call_and_handle_interrupt(
File "/opt/conda/lib/python3.10/site-packages/lightning/pytorch/trainer/call.py", line 44, in _call_and_handle_interrupt
return trainer_fn(*args, **kwargs)
File "/opt/conda/lib/python3.10/site-packages/lightning/pytorch/trainer/trainer.py", line 579, in _fit_impl
self._run(model, ckpt_path=ckpt_path)
File "/opt/conda/lib/python3.10/site-packages/lightning/pytorch/trainer/trainer.py", line 986, in _run
results = self._run_stage()
File "/opt/conda/lib/python3.10/site-packages/lightning/pytorch/trainer/trainer.py", line 1032, in _run_stage
self.fit_loop.run()
File "/opt/conda/lib/python3.10/site-packages/lightning/pytorch/loops/fit_loop.py", line 205, in run
self.advance()
File "/opt/conda/lib/python3.10/site-packages/lightning/pytorch/loops/fit_loop.py", line 363, in advance
self.epoch_loop.run(self._data_fetcher)
File "/opt/conda/lib/python3.10/site-packages/lightning/pytorch/loops/training_epoch_loop.py", line 138, in run
self.advance(data_fetcher)
File "/opt/conda/lib/python3.10/site-packages/lightning/pytorch/loops/training_epoch_loop.py", line 204, in advance
batch, _, __ = next(data_fetcher)
File "/opt/conda/lib/python3.10/site-packages/lightning/pytorch/loops/fetchers.py", line 133, in __next__
batch = super().__next__()
File "/opt/conda/lib/python3.10/site-packages/lightning/pytorch/loops/fetchers.py", line 60, in __next__
batch = next(self.iterator)
File "/opt/conda/lib/python3.10/site-packages/lightning/pytorch/utilities/combined_loader.py", line 341, in __next__
out = next(self._iterator)
File "/opt/conda/lib/python3.10/site-packages/lightning/pytorch/utilities/combined_loader.py", line 78, in __next__
out[i] = next(self.iterators[i])
File "/opt/conda/lib/python3.10/site-packages/torch/utils/data/dataloader.py", line 628, in __next__
data = self._next_data()
File "/opt/conda/lib/python3.10/site-packages/torch/utils/data/dataloader.py", line 671, in _next_data
data = self._dataset_fetcher.fetch(index) # may raise StopIteration
File "/opt/conda/lib/python3.10/site-packages/torch/utils/data/_utils/fetch.py", line 58, in fetch
data = [self.dataset[idx] for idx in possibly_batched_index]
File "/opt/conda/lib/python3.10/site-packages/torch/utils/data/_utils/fetch.py", line 58, in <listcomp>
data = [self.dataset[idx] for idx in possibly_batched_index]
File "/home/jovyan/trv/MVP_SPACE/mvp_space/src/train_pipeline/dataset.py", line 184, in __getitem__
sample = self.transform(
File "/opt/conda/lib/python3.10/site-packages/albumentations/core/composition.py", line 493, in __call__
self.preprocess(data)
File "/opt/conda/lib/python3.10/site-packages/albumentations/core/composition.py", line 533, in preprocess
self._preprocess_processors(data)
File "/opt/conda/lib/python3.10/site-packages/albumentations/core/composition.py", line 560, in _preprocess_processors
processor.preprocess(data)
File "/opt/conda/lib/python3.10/site-packages/albumentations/core/utils.py", line 155, in preprocess
data[data_name] = self.check_and_convert(data[data_name], shape, direction="to")
File "/opt/conda/lib/python3.10/site-packages/albumentations/core/bbox_utils.py", line 306, in check_and_convert
self.check(converted_data, shape)
File "/opt/conda/lib/python3.10/site-packages/albumentations/core/bbox_utils.py", line 314, in check
check_bboxes(data)
File "/opt/conda/lib/python3.10/site-packages/albumentations/augmentations/utils.py", line 189, in wrapper
return func(*args, **kwargs)
File "/opt/conda/lib/python3.10/site-packages/albumentations/core/bbox_utils.py", line 549, in check_bboxes
raise ValueError(
ValueError: Expected x_min for bbox [ 4.372159 20.666666 4.9176135 26.666666 1. ] to be in the range [0.0, 1.0], got 4.372159004211426.
Epoch 0: 0%| | 0/569 [00:00<?, ?it/s]
However, pascal_voc is NOT supposed to be between [0, 1]... What is going on here?
Something more weird are the numbers, they are all integers, as you can see by this assert I added at the start. I don't know when they are transformed to float.
for bbox in torch_target["boxes"]:
assert np.all([int(x) == x for x in bbox])
if self.transform:
sample = self.transform(
image=torch_img.numpy(),
bboxes=torch_target["boxes"],
class_labels=torch_target["labels"],
)
I am using albumentations for my object detection data augmentation because it deals with bbox at the same time. My format is [xmin, ymin, xmax, ymax] with pascal_voc format.
A.Compose(
[
A.HorizontalFlip(p=0.5),
A.VerticalFlip(p=0.5),
A.OneOf(
[
A.ShiftScaleRotate(
shift_limit=0.2,
scale_limit=(-0.1, 0.6),
rotate_limit=0,
p=0.5,
),
A.RandomSizedCrop(
min_max_height=[200, 352], size=(352, 2048), p=0.5
),
],
p=0.5,
),
A.AdditiveNoise(
noise_type="uniform", noise_params={"ranges":[(-0.2, 0.2)]}, p=0.5
),
ToTensorV2(),
],
bbox_params=A.BboxParams(
format="pascal_voc", label_fields=["class_labels"]
),
)
Then I create my data as:
def __getitem__(self, index):
data = get_data(index)
torch_target = _to_torch_target(data)
torch_img = _get_torch_image(data)
if self.transform:
sample = self.transform(
image=torch_img.numpy(),
bboxes=torch_target["boxes"],
class_labels=torch_target["labels"],
)
torch_img = sample["image"]
torch_target["boxes"] = torch.Tensor(sample["bboxes"])
torch_target["labels"] = torch.Tensor(sample["class_labels"])
torch_target["boxes"] = torch.as_tensor(torch_target["boxes"], dtype=torch.float32)
torch_target["labels"] = torch.as_tensor(torch_target["labels"], dtype=torch.int64)
return torch_img, torch_target
But I am getting an error:
Traceback (most recent call last):
File "/opt/conda/lib/python3.10/runpy.py", line 196, in _run_module_as_main
return _run_code(code, main_globals, None,
File "/opt/conda/lib/python3.10/runpy.py", line 86, in _run_code
exec(code, run_globals)
File "/home/.../main.py", line 116, in <module>
train(results_path, dataset_path, weights_path, **conf)
File "/home/.../main.py", line 77, in train
trainer.fit(model, train_dataloader, val_dataloader)
File "/opt/conda/lib/python3.10/site-packages/lightning/pytorch/trainer/trainer.py", line 543, in fit
call._call_and_handle_interrupt(
File "/opt/conda/lib/python3.10/site-packages/lightning/pytorch/trainer/call.py", line 44, in _call_and_handle_interrupt
return trainer_fn(*args, **kwargs)
File "/opt/conda/lib/python3.10/site-packages/lightning/pytorch/trainer/trainer.py", line 579, in _fit_impl
self._run(model, ckpt_path=ckpt_path)
File "/opt/conda/lib/python3.10/site-packages/lightning/pytorch/trainer/trainer.py", line 986, in _run
results = self._run_stage()
File "/opt/conda/lib/python3.10/site-packages/lightning/pytorch/trainer/trainer.py", line 1032, in _run_stage
self.fit_loop.run()
File "/opt/conda/lib/python3.10/site-packages/lightning/pytorch/loops/fit_loop.py", line 205, in run
self.advance()
File "/opt/conda/lib/python3.10/site-packages/lightning/pytorch/loops/fit_loop.py", line 363, in advance
self.epoch_loop.run(self._data_fetcher)
File "/opt/conda/lib/python3.10/site-packages/lightning/pytorch/loops/training_epoch_loop.py", line 138, in run
self.advance(data_fetcher)
File "/opt/conda/lib/python3.10/site-packages/lightning/pytorch/loops/training_epoch_loop.py", line 204, in advance
batch, _, __ = next(data_fetcher)
File "/opt/conda/lib/python3.10/site-packages/lightning/pytorch/loops/fetchers.py", line 133, in __next__
batch = super().__next__()
File "/opt/conda/lib/python3.10/site-packages/lightning/pytorch/loops/fetchers.py", line 60, in __next__
batch = next(self.iterator)
File "/opt/conda/lib/python3.10/site-packages/lightning/pytorch/utilities/combined_loader.py", line 341, in __next__
out = next(self._iterator)
File "/opt/conda/lib/python3.10/site-packages/lightning/pytorch/utilities/combined_loader.py", line 78, in __next__
out[i] = next(self.iterators[i])
File "/opt/conda/lib/python3.10/site-packages/torch/utils/data/dataloader.py", line 628, in __next__
data = self._next_data()
File "/opt/conda/lib/python3.10/site-packages/torch/utils/data/dataloader.py", line 671, in _next_data
data = self._dataset_fetcher.fetch(index) # may raise StopIteration
File "/opt/conda/lib/python3.10/site-packages/torch/utils/data/_utils/fetch.py", line 58, in fetch
data = [self.dataset[idx] for idx in possibly_batched_index]
File "/opt/conda/lib/python3.10/site-packages/torch/utils/data/_utils/fetch.py", line 58, in <listcomp>
data = [self.dataset[idx] for idx in possibly_batched_index]
File "/home/jovyan/trv/MVP_SPACE/mvp_space/src/train_pipeline/dataset.py", line 184, in __getitem__
sample = self.transform(
File "/opt/conda/lib/python3.10/site-packages/albumentations/core/composition.py", line 493, in __call__
self.preprocess(data)
File "/opt/conda/lib/python3.10/site-packages/albumentations/core/composition.py", line 533, in preprocess
self._preprocess_processors(data)
File "/opt/conda/lib/python3.10/site-packages/albumentations/core/composition.py", line 560, in _preprocess_processors
processor.preprocess(data)
File "/opt/conda/lib/python3.10/site-packages/albumentations/core/utils.py", line 155, in preprocess
data[data_name] = self.check_and_convert(data[data_name], shape, direction="to")
File "/opt/conda/lib/python3.10/site-packages/albumentations/core/bbox_utils.py", line 306, in check_and_convert
self.check(converted_data, shape)
File "/opt/conda/lib/python3.10/site-packages/albumentations/core/bbox_utils.py", line 314, in check
check_bboxes(data)
File "/opt/conda/lib/python3.10/site-packages/albumentations/augmentations/utils.py", line 189, in wrapper
return func(*args, **kwargs)
File "/opt/conda/lib/python3.10/site-packages/albumentations/core/bbox_utils.py", line 549, in check_bboxes
raise ValueError(
ValueError: Expected x_min for bbox [ 4.372159 20.666666 4.9176135 26.666666 1. ] to be in the range [0.0, 1.0], got 4.372159004211426.
Epoch 0: 0%| | 0/569 [00:00<?, ?it/s]
However, pascal_voc is NOT supposed to be between [0, 1]... What is going on here?
Something more weird are the numbers, they are all integers, as you can see by this assert I added at the start. I don't know when they are transformed to float.
for bbox in torch_target["boxes"]:
assert np.all([int(x) == x for x in bbox])
if self.transform:
sample = self.transform(
image=torch_img.numpy(),
bboxes=torch_target["boxes"],
class_labels=torch_target["labels"],
)
1 Answer
Reset to default 0So, it was the fact that I was already using a torch format numpy image. That has channels first. The solution was using
image=torch_img.permute(1, 2, 0).numpy()
发布者：admin，转转请注明出处：http://www.yc00.com/questions/1744971427a4603988.html
评论列表（0条）