Skip to content
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

add RaDeGS #317

Open
wants to merge 2 commits into
base: main
Choose a base branch
from
Open

add RaDeGS #317

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
3c9dc0e
add RaDeGS
BaowenZ Aug 7, 2024
eb150f8
fix persp_proj_vjp
BaowenZ Aug 8, 2024
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
26 changes: 22 additions & 4 deletions examples/datasets/colmap.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -138,9 +138,14 @@ def __init__(
image_dir_suffix = ""
colmap_image_dir = os.path.join(data_dir, "images")
image_dir = os.path.join(data_dir, "images" + image_dir_suffix)
for d in [image_dir, colmap_image_dir]:
if not os.path.exists(d):
raise ValueError(f"Image folder {d} does not exist.")

if not os.path.exists(colmap_image_dir):
raise ValueError(f"Image folder {colmap_image_dir} does not exist.")
if not os.path.exists(image_dir):
image_dir = colmap_image_dir
self.factor = factor
else:
self.factor = None

# Downsampled images may have different names vs images used for COLMAP,
# so we need to map between the two sorted lists of files.
Expand Down Expand Up @@ -249,12 +254,22 @@ def __len__(self):

def __getitem__(self, item: int) -> Dict[str, Any]:
index = self.indices[item]
image = imageio.imread(self.parser.image_paths[index])[..., :3]
image = imageio.imread(self.parser.image_paths[index])
if image.shape[-1]==4:
mask = image[...,-1]
Copy link

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Should the conversion to a binary mask be done here? (you're doing data["mask"]/255<0.5 later on)

image = image[...,:3]
else:
mask = None
camera_id = self.parser.camera_ids[index]
K = self.parser.Ks_dict[camera_id].copy() # undistorted K
params = self.parser.params_dict[camera_id]
camtoworlds = self.parser.camtoworlds[index]

if self.parser.factor is not None:
image = cv2.resize(image,self.parser.imsize_dict[camera_id])
if mask is not None:
mask = cv2.resize(mask,self.parser.imsize_dict[camera_id])

if len(params) > 0:
# Images are distorted. Undistort them.
mapx, mapy = (
Expand All @@ -274,11 +289,14 @@ def __getitem__(self, item: int) -> Dict[str, Any]:
K[0, 2] -= x
K[1, 2] -= y



data = {
"K": torch.from_numpy(K).float(),
"camtoworld": torch.from_numpy(camtoworlds).float(),
"image": torch.from_numpy(image).float(),
"image_id": item, # the index of the image in the dataset
"mask": torch.from_numpy(mask).float(),
}

if self.load_depths:
Expand Down
Loading