Mitigation of attitude and gyro errors through vision aiding
Bibliographic record
Abstract
Accurate positioning of first responders, electronic monitoring, and military personnel is often critical in GNSS denied environments. In such environments, inertial navigation systems (INS) are typically the preferred tool to be used for navigation. However, the gyros suffer from errors including biases, scale factors and g-dependent errors being the most significant ones. In order to sustain an accurate navigation solution for long durations, the gyroscope errors have to be measured and mitigated. Ideally, this calibration is done in situ. The attitude obtained using visual information is independent of the errors affecting the gyroscope. Human-made environments are commonly full of straight and parallel lines found in orthogonal directions. Perspective projection mapping transforms three-dimensional scenes into two-dimensional images. The process maintains the straight lines but modifies their parallelism resulting in an apparent point intersection of the lines. This point is called the vanishing point. Lines in three orthogonal directions constitute three vanishing points. The vanishing point locations are dependent on the camera rotation, but not camera translation. By monitoring the motion of the vanishing point locations in consecutive images, the relative roll, heading and pitch attitudes may be obtained. The absolute attitude, known from some a priori knowledge of the building layout, is then used to update the inertial navigation filter. Over time the visual measurements mitigate the cumulative errors of the gyro bias, scale factor and g-dependent bias. The performance of the vision-aided INS based navigation approach is evaluated herein. A camera is attached to a backpack and foot of a user moving through typical pedestrian based environments. The case of a foot-mounted camera is unique because of the high accelerations experienced during the human gait. The visual-aiding correction is found to significantly improve the attitude accuracy, especially the heading. Using the body solution, namely the camera and INS attached to a backpack, the vision-aiding yielded a 93 % improvement in the heading error during evaluation tests. With a foot-mounted solution, namely the INS and camera attached to the ankle of the user, the horizontal position error decreased by 34 %.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".