Measuring visual attention to online videos with a mouse cursor window paradigm: considerations for large scale data collection
Bibliographic record
Abstract
Previous research has allowed for large scale datasets to be created from mouse-based measures of attention to images (e.g., SALICON). However, there has not been a similar, easy-to-use solution for large scale data collection of attention to video stimuli. Here, we demonstrate our novel mouse-based measure of attention that can be used with videos in online experiments. Our results show similar performance between this paradigm and eye tracking in terms of the attended regions of interest in video. This paradigm tracks the user’s mouse location as they use their mouse to move a window of high resolution around an otherwise blurred screen. To view video content in more detail, the user moves their mouse window to that location. This results in a robust measure of visual attention that can be used to identify regions of video content participants find most salient or informative. Our research has compared eye movements from the DIEM dataset to mouse movements collected from online participants watching videos with a mouse-contingent bi-resolution display. To test the settings of the mouse-based method, participants experienced large, medium, and small window sizes and blur levels, in a 3x3 within-subjects factorial design. New results show that the motions made between these methodologies differ, but they result in visits to similar regions of interest in video. This suggests that mouse-based methods may not replicate effects that require the speed and ease of eye movements, but can replicate effects regarding attention to salient, informative, and preferred regions of content. These findings further support the validity of this method in large scale data collection on users’ attention to video stimuli–valuable to both experimental research, and the training and testing of video saliency models. With this presentation, we will discuss plans to make our methodology available for use with online experiment software.
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 | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.000 | 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".