Segmentation mechanisms are sensitive to and can segment by higher-order statistics in naturalistic textures
Bibliographic record
Abstract
Texture segmentation depends on the statistical properties of the textures in question, but which properties are biologically relevant remains unclear. Previously, we determined that contrast boundary segmentation in a single texture was affected by global phase structure – in particular, the texture density (VSS 2009). Here, we manipulate density and broadband phase alignment for boundaries between pairs of textures differing in orientation or phase alignment to uncover the statistical sensitivities of segmentation mechanisms. We created synthetic micropattern textures that mimic important statistical properties of natural textures (VSS 2009). We were able to remove all higher-order statistics by globally phase-scrambling the texture, or remove local phase alignments by phase-scrambling the micropatterns, while varying texture sparseness by changing the number of micropatterns. We created two types of texture boundaries using a quilting method: (1) orientation modulations, where one texture had vertical micropatterns and the other horizontal, and (2) phase alignment boundaries between different pairings of the intact, local (LS), and global (GS) scramble conditions. We obtained modulation-depth thresholds for all boundary types at a series of micropattern densities. Orientation-defined boundaries become easier to segment as density increases, with boundaries between GS textures being easier to segment than those between either intact or LS textures. As density increases, boundaries between GS and either intact or LS become more difficult, but boundaries between intact and LS textures become easier. We observe that boundaries defined by changes in phase alignment are no more difficult to segment than those defined by changes in orientation These results lend support to the idea that sparseness is an important texture dimension impacting performance in segmentation tasks. Our findings suggest that early inputs to segmentation mechanisms are sensitive to higher-order statistics such as sparseness as well as simple attributes such as contrast and orientation.
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 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 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".