Evidence of lack of integration of binocular disparity and motion parallax in object segmentation
Bibliographic record
Abstract
Rogers and Graham (1979) showed that motion parallax can support depth percepts comparable to those generated by stereopsis. However, several experiments with physical stimuli have shown that when both are present in depth estimation tasks, observers appear to rely on stereoscopic information. One explanation is that these tasks bias observers towards using stereopsis. Here we devised a novel segmentation task to evaluate cue integration that should benefit from relative motion. Observers viewed two superimposed, frontoparallel, horizontal wavy lines using a virtual reality headset. On each side of the set of lines, a probe was aligned with the end of one of the curves; observers indicated whether the two probes were coincident with the same curve. We generated two levels of complexity by manipulating the curvature. Using the method of constant stimuli we varied the depth separation between the two curves from 0 to 2.4cm and assessed performance using stereopsis alone, motion parallax alone (monocular) or both cues present. On monocular trials, observers moved their heads laterally by 6cm. As anticipated, accuracy was lowest when there was no depth offset between the two curves. The low complexity condition was generally easy, and performance was the same across conditions. However, in our high complexity condition, performance was near chance when there was no depth separation and gradually increased as the depth offset increased. Accuracy was similar in the stereopsis only and combined conditions but significantly poorer when only motion parallax was available. This was true even when the range of head motion was doubled. In sum, despite using a task that should benefit from relative motion and the fact that the two depth cues provided consistent information, our results echo previous studies in showing an apparent lack of integration of motion parallax and stereopsis.
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 | 0.001 | 0.006 |
| Meta-epidemiology (narrow) | 0.000 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.002 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 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".