Skip to content

⚡️ Speed up function manage_crops_metadata by 274% in PR #1654 (fix/stitch-block-image-dimensions) #1655

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

Closed
wants to merge 1 commit into from
Closed

⚡️ Speed up function manage_crops_metadata by 274% in PR #1654 (fix/stitch-block-image-dimensions) #1655

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
14 changes: 12 additions & 2 deletions inference/core/workflows/core_steps/fusion/detections_stitch/v1.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -215,8 +215,18 @@ def manage_crops_metadata(
detections.data[PARENT_COORDINATES_KEY] -= offset
if ROOT_PARENT_COORDINATES_KEY in detections.data:
detections.data[ROOT_PARENT_COORDINATES_KEY] -= offset
detections.data[IMAGE_DIMENSIONS_KEY] = np.array([dimensions] * len(detections))
detections.data[PARENT_ID_KEY] = np.array([parent_id] * len(detections))

# PRE-ALLOCATE ARRAYS for efficiency
num_detections = len(detections)
# np.full avoids creating a list only to pass it to np.array, which is costly for large num_detections
detections.data[IMAGE_DIMENSIONS_KEY] = np.full(
(num_detections, 2), dimensions, dtype=np.int32
)
# Use object dtype for non-numeric data, fill directly rather than creating intermediate list
arr_parent_ids = np.empty(num_detections, dtype=object)
arr_parent_ids.fill(parent_id)
detections.data[PARENT_ID_KEY] = arr_parent_ids

return detections


Expand Down