Psychophysical evaluation of planar shape representations for object recognition
Bibliographic record
Abstract
Intermediate areas of the object pathway appear to represent shape in terms of features of moderate complexity, however the precise nature of this distributed code remains unclear. Here we use a novel method to evaluate the efficiency with which three candidate representations (Fourier Descriptors, Shapelets and Formlets) capture the planar shape information required for humans to reliably recognize objects. The Fourier Descriptor representation is the Fourier transform of the points defining the object boundary, represented as complex numbers; a good approximation to a shape is attained by truncating this Fourier sequence. Shapelets are a wavelet version of Fourier Descriptors, where each component is localized in both frequency and position along the curve, and these are computed by matching pursuit. Formlets represent shape as a series of smooth localized deformations applied to an embryonic shape (an ellipse in our case), also computed using matching pursuit. We employed a database of 77 animal shapes from 11 categories. In objective terms (Euclidean error), these shapes are most efficiently coded by Shapelets, followed by Fourier Descriptors, and finally Formlets. To evaluate subjective efficiency, shapes were rendered using each of these three representations; the observer’s task was to identify the category of each shape from four alternatives. For each representation, the number of shape components ranged from 1 to 10; a representation that reaches subjective threshold with fewer components may be closer to the code employed by the human visual system. For all 6 observers, Shapelets were found to have lowest threshold (mean of 1.8±0.4 components), followed by Fourier Descriptors (4.0±0.4 components), and finally Formlets (5.5±0.6 components). Interestingly, however, both Shapelets and Formlets reach subjective thresholds at a higher mean objective error than Fourier Descriptors, suggesting that the human visual system relies upon localized basis functions for shape representation. Meeting abstract presented at VSS 2015
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 | 0.003 | 0.020 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.000 | 0.001 |
| Research integrity | 0.001 | 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".