Not Losing Your Way: Mitigating Visual-Inertial Odometry Drift through On-Demand Visual Feature Creation
Not Losing Your Way: Mitigating Visual-Inertial Odometry Drift through On-Demand Visual Feature Creation
This research proposes an active environment-modification framework for UAVs that mitigates visual-inertial odometry (VIO) drift in repetitive or poorly textured environments by creating visual features on demand.
The study presents a simulation and experimental evaluation of an active feature creation framework for UAVs that reduces VIO position drift by approximately 35% in repetitive environments, though results are limited to specific simulated corridors and a 900-run factorial campaign without broader real-world deployment validation.
Source: arXiv Robotics — research abstracts · Read original article ↗
Loading article text…
What the source reports
Publisher-reported claims, with original evidence. These results have not been independently verified by RoboSignal.
- Environment
- SimulationOpen source S3
Reported numbers
runs
900
Reported trials
View original evidence
ates visual features on demand by releasing ground markers or marking walls through physical contact. An information-based indicator triggers interventions using temporal confirmation and hysteresis. Evaluation includes three simulated environments and a 900-run factorial campaign. Lower cruise speed and active marking are associated with reduced divergence, while accuracy among non-divergent runs depends on speed-st
Open source S3runs
5
Reported trials
View original evidence
rategy interactions. Corridor experiments, with five runs per configuration, use OptiTrack for flight control while VIO-based marking decisions remain active online. The combined framework reduces the reported position root mean square error by approximately 35% relative to baseline VIO, supporting active feature creation for drift mitigation.
Open source S4
- paper: not_reported
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
ates visual features on demand by releasing ground markers or marking walls through physical contact. An information-based indicator triggers interventions using temporal confirmation and hysteresis. Evaluation includes three simulated environments and a 900-run factorial campaign. Lower cruise speed and active marking are associated with reduced divergence, while accuracy among non-divergent runs depends on speed-st
Open source S3
rategy interactions. Corridor experiments, with five runs per configuration, use OptiTrack for flight control while VIO-based marking decisions remain active online. The combined framework reduces the reported position root mean square error by approximately 35% relative to baseline VIO, supporting active feature creation for drift mitigation.
Open source S4
Source：arXiv Robotics — research abstracts · arxiv.org