Real-world visual search goes beyond eye movements: Active searchers select 3D scene viewpoints too
Bibliographic record
Abstract
Visual search is a ubiquitous task; people search for objects on a daily basis. However, the majority of the existing visual search literature focuses on passive search on a 2D computer screen, a far cry from emulating a real-world environment. Search is a real-world task that involves active observation. Search targets may be occluded, completely out of the observer's line of sight, or oriented in unconventional ways. This is typically mitigated by actively selecting viewpoints, an important aspect of search behaviour with limited scope on a computer screen. Our goal was to explore viewpoint selection in active visual search. Subject eye and head movements were tracked as they moved freely while searching for toy objects in a controlled 3-dimensional environment, yielding the first such record of search-driven viewpoint selection. We found that subjects utilized their full range of eye and head motion to move from viewpoint to viewpoint, apparently employing a variety of objectives including changing viewing height and pose depending on object 3D pose. Subjects were also adept at selecting unobstructed views to search through otherwise occluded areas with objects. Furthermore, subjects completed the search task with high accuracy, even with no training on the environment. Although no learning was found in terms of accuracy over the duration of the experiment, increases in efficiency were found for other metrics such as response time, number of fixations, and distance travelled, particularly in target present trials where the target was not visible from the starting location. These results paint the story of a visual system that selects and moves to useful and informative views to facilitate the successful execution of an active visual search task, and stresses the significance of active vision research in understanding how vision is used in naturalistic environments.
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 | 0.002 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| 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".