Fix off-by-one crash in forward_sliding1 for T ≡ 1 (mod chunk_size)#6
Merged
Merged
Conversation
The per-chunk loop iterates `num_chunks = ceil((T-1)/chunk_size)` times (driven by the pairwise frame count B_pair = T-1), but points / masks / world_points / camera_poses contain T frames and are split into `ceil(T/chunk_size)` chunks. When T ≡ 1 (mod chunk_size) — e.g. T=133 with chunk_size=12 — the last chunk is never appended, so the returned points sequence is one frame short. This surfaces downstream as a RuntimeError in infer_pair: "The size of tensor a (132) must match the size of tensor b (131)" at `flow3d = flows3d_e[None] + points[None, 0:-1]`, and also in get_aligned_scene_flow_temporal. Fix: append the leftover trailing chunks after the loop so points/masks/ world_points/camera_poses regain their full T frames. Verified by rerunning a 133-frame clip in 3d_efep mode (previously crashed, now completes).
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
The per-chunk loop iterates
num_chunks = ceil((T-1)/chunk_size)times (driven by the pairwise frame count B_pair = T-1), but points / masks / world_points / camera_poses contain T frames and are split intoceil(T/chunk_size)chunks. When T ≡ 1 (mod chunk_size) — e.g. T=133 with chunk_size=12 — the last chunk is never appended, so the returned points sequence is one frame short.This surfaces downstream as a RuntimeError in infer_pair:
"The size of tensor a (132) must match the size of tensor b (131)"
at
flow3d = flows3d_e[None] + points[None, 0:-1], and also in get_aligned_scene_flow_temporal.Fix: append the leftover trailing chunks after the loop so points/masks/ world_points/camera_poses regain their full T frames. Verified by rerunning a 133-frame clip in 3d_efep mode (previously crashed, now completes).