VIOVIO Survey BenchmarkGitHub Pages reproducible benchmark

Evaluation protocol

Rules before numbers

The benchmark is evaluated with EPICA / EPA and OpenVINS-style documentation discipline. The protocol defines valid trajectories, metric semantics, missing-data rules, and artifact requirements.

Protocol

Public evaluation rules

These rules are intentionally short and strict. They prevent the website from becoming a leaderboard with unclear inputs.

01

Primary trajectories are current-timestep odometry estimates.

02

Systems with SLAM components must be configured for odometry-only reporting in this benchmark.

03

EPA is the primary evaluation pipeline for alignment, association, metric generation, plots, summaries, and paper tables.

04

OpenVINS ov_eval-style commands are compatibility references, not the primary publication path.

05

TBD, N/A, Partial, and Unresolved rows are visible but excluded from numeric averages unless a visible rule states otherwise.

06

Every public result row must carry system, dataset, sequence, mode, command source, artifact path, and missing-data status.

Metrics

Metric definitions and missing-data policy

The dashboard reports metric families only when their EPA or profiling artifacts are present.

Metric policy

MetricNameScopeMissing-data rule
ATEAbsolute Trajectory ErrorFull trajectory after configured EPA alignmentTBD until metrics.json or summary.csv exists
RPERelative Pose ErrorLocal drift over configured distance, time, or frame pairsTBD unless pair policy is recorded
DriftDistance Drift RateTranslation RPE normalized by path-length segmentTBD when segment policy is not available
SuccessDrift-valid Success RateReference-path percentage satisfying the configured criterionTBD unless EPA drift-valid output is present
RuntimeRuntimeTracking, estimation, and total per-frame timingTBD until timing CSV is recorded
ResourceCPU and MemoryProcess-level resource monitor logs on a labeled platformTBD until monitor log is recorded