Skip to content

Fix height/width swap and other bugs in dataset utils #910

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
Mr-Neutr0n wants to merge 1 commit into
base: main
Choose a base branch
from
Open

Fix height/width swap and other bugs in dataset utils #910

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
8 changes: 4 additions & 4 deletions opensora/datasets/utils.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -76,7 +76,7 @@ def download_url(input_path):
base_name = os.path.basename(input_path)
output_path = os.path.join(output_dir, base_name)
img_data = requests.get(input_path).content
with open(output_path, "wb", encoding="utf-8") as handler:
with open(output_path, "wb") as handler:
handler.write(img_data)
print(f"URL {input_path} downloaded to {output_path}")
return output_path
Expand Down Expand Up @@ -266,7 +266,7 @@ def rand_size_crop_arr(pil_image, image_size):

# get random start pos
h_start = random.randint(0, max(len(arr) - height, 0))
w_start = random.randint(0, max(len(arr[0]) - height, 0))
w_start = random.randint(0, max(len(arr[0]) - width, 0))

# crop
return Image.fromarray(arr[h_start : h_start + height, w_start : w_start + width])
Expand Down Expand Up @@ -356,7 +356,7 @@ def rescale_image_by_path(path: str, height: int, width: int):
raise ValueError("The image is invalid or empty.")

# resize image
resize_transform = transforms.Resize((width, height))
resize_transform = transforms.Resize((height, width))
resized_image = resize_transform(image)

# save resized image back to the original path
Expand Down Expand Up @@ -384,7 +384,7 @@ def rescale_video_by_path(path: str, height: int, width: int):
raise ValueError("The video is invalid or empty.")

# Resize video frames using a performant method
resize_transform = transforms.Compose([transforms.Resize((width, height))])
resize_transform = transforms.Compose([transforms.Resize((height, width))])
resized_video = torch.stack([resize_transform(frame) for frame in video])

# Save resized video back to the original path
Expand Down