Natural size-distance scaling reduces, but does not eliminate, depth matching errors from conflicting occlusion and stereopsis
Bibliographic record
Abstract
Stereoscopic depth matching is significantly degraded when occlusion information conflicts with depth from binocular disparity. However, in these studies the visual angle of the target was held constant while in natural viewing image size changes linearly with object distance. Thus, it is not clear whether the disruption in performance can be solely attributed to the discrepant occlusion and disparity signals. Here we evaluated the combined effects of size, occlusion and binocular disparity using a depth matching paradigm in mixed-reality. The virtual stimulus was a green letter ‘A’ presented using an augmented reality display. It was superimposed on a physical surface with variable transparency fixed at 1.2 m. The target letter was placed at one of eight distances (0.9 - 1.6 m), including the surface location. The letter was rendered either with a fixed size (variable retinal angle) or with size scaled to maintain a constant visual angle. Observers matched the distance of a virtual probe to the perceived distance of the letter in three conditions where the surface was: opaque, transparent or absent. We found that depth matches were accurate and there was no effect of size scaling in both the ‘no surface’ and ‘transparent surface’ conditions. This was also the case when the letter appeared in front of the surface. However, when the letter was positioned beyond the opaque surface (maximum cue conflict) its position was systematically underestimated. Introducing correct size scaling reduced this error but did not eliminate it. In sum, when occlusion and disparity are in conflict our results show that using a fixed retinal size exacerbates depth matching errors. When retinal size varies (as in the natural world) depth matching is more accurate but significant underestimates remain. When occlusion and disparity signals are in agreement, size has little impact on performance.
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 | 0.000 | 0.002 |
| 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 | 0.001 |
| Open science | 0.000 | 0.001 |
| Research integrity | 0.000 | 0.000 |
| 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".