The impact of conflicting ordinal and metric depth information on depth matching
Bibliographic record
Abstract
Under natural viewing conditions binocular disparity can provide metric depth information; many of the monocular depth cues, such as occlusion, provide depth order only. Nonetheless, when put in conflict there is evidence that occlusion can influence the direction and magnitude of perceived depth from stereopsis. Here we explored the integration of depth information from occlusion and binocular disparity in complex real-world environments using a depth matching paradigm. The virtual stimulus was a green letter ‘A’ presented using a Microsoft HoloLens augmented reality (AR) display and superimposed on a real frontoparallel surface at 1.2 m. The letter was placed at one of eight positions – between 0.9 and 1.6 m, including the surface location. Observers matched the distance of a probe to the perceived distance of the letter by moving it with a sliding pole. For comparison, observers performed the same task without the physical surface. Our results show that when the surface was absent or the letter was rendered in front of the surface the letter was accurately localized. However, when the letter was rendered beyond the surface, observers progressively underestimated the letter’s distance, even though the relative disparity between the probe and the target should have been equally informative at all locations. This pattern of results suggests that 1) observers are unable to ignore conflicts between occlusion and binocular disparity and 2) the occlusion conflict biases the perceived position of the target in the direction of the occluder. Our results are well modelled using a Bayesian ideal observer with an asymmetric likelihood for an occlusion cue representing letter positions in front of vs beyond the surface. In addition to providing insight into the integration of ordinal and metric depth information, these results speak to the impact of such errors in AR on user interactions.
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 machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.025 |
| Meta-epidemiology (narrow) | 0.000 | 0.001 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.002 | 0.002 |
| Open science | 0.001 | 0.003 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 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 source (direct Gemma or distilled Codex), 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".