cuvslam: multicam mode — Spot's five-camera rig - #3485
Draft
jeff-hykin wants to merge 9 commits into
Draft
Conversation
Multisensor mode with depth anchoring scale from the side/back cameras; front pair contributes features only (its floor-dominated depth degrades ATE 1.7m -> 14.5m on spot_small_loop). Portrait front images are fed as their landscape rotation with adjusted intrinsics/extrinsics.
4s of steady walking (all five cameras + depth + odometry + tf) cut from spot_small_loop.db; the test replays it through the Multisensor rig and sanity-checks pose flow against the recorded odometry.
camera_mode=multicam with depth_camera_frames runs the dimSLAM binary in cuVSLAM's Multisensor mode: an explicit rig of mono cameras with metric scale anchored by the named cameras' registered depth. Pin the dimSLAM build to the multicam rev.
❌ 5 Tests Failed:
View the top 3 failed test(s) by shortest run time
View the full list of 2 ❄️ flaky test(s)
To view more test analytics, go to the Test Analytics Dashboard |
…-rc1 mypy override + importorskip cover machines where the pycuvslam wheel is not installed. The test stays on the GPU deliberately: Multisensor's depth path runs CUDA kernels even in an ENFORCE_GPU=OFF build, so CPU mode aborts with CUDA error 700 and cannot serve as the reproducible path.
…t/cuvslam_multi # Conflicts: # dimos/mapping/cuvslam/cuvslam.py
…cuvslam The test now boots the dimSLAM binary via CuvslamOdometry (multicam, three landscape cameras + depth) and feeds the LFS snippet over LCM, asserting the published odometry flows and agrees with the robot's own. The pycuvslam replay demo moves out of the repo; the PR ships only the native path.
autoconnect(SpotSnippetPublisher, CuvslamOdometry, OdometrySink): the publisher module replays the LFS snippet on its Out ports, the tracker runs the dimSLAM binary, the sink collects odometry — no hand-wired transports in the test body.
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
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.
Multi-camera cuVSLAM, tested on the Spot data