Can a saliency model using feature sets derived cityscapes predict cultural differences in visual search asymmetry?
Bibliographic record
Abstract
Visual search asymmetries in line length (long or short), line orientation (tilted or vertical), and circle with line versus circle vary in presence and magnitude depending on where they are measured (e.g., Canada, Japan, and the United States, see Ueda et al., 2018). These results indicate that our visual cognition is affected by the environments around us. What induces these differences? Several previous studies suggest that it may due to the cityscape and orthographies of each location. It has been shown that visual attention and eye movements change after viewing Japanese and American cityscapes for a certain amount of time (Miyamoto et al. 2006; Ueda & Komiya, 2012). A study dealing directly with search asymmetry showed that visual saliency calculated using the Attention based on Information Maximization (AIM; Bruce & Tsotsos, 2009) model, which derives a set of visual features from orthographic characters (alphabets, Japanese hiragana, and Kanji characters) produced different saliency levels in line length asymmetry (Saiki, 2020). In this study, we used the AIM model to investigate whether differences in the scenery of different locations can produce different visual saliency in the search asymmetry. In the original AIM model, visual feature sets were derived from various scene images, while instead of this, we collected more than 10,000 photos of cities across Japan, and used only them to derive visual feature sets with independent component analysis (ICA). Comparing visual saliency calculated by visual features derived from the original image set with the image set of Japanese cities, consistent results across image sets were obtained for search asymmetry. These results suggest that the AIM model with feature sets derived from scenery can robustly predict visual search asymmetry, and cultural differences in search asymmetry may emerge not from the cityscapes in each culture, but rather from long-term experiences with orthographic letters.
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 distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.000 | 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 teacher head, 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".