Multi-Class Gaze Detection in a Dynamic Environment
Bibliographic record
Abstract
Developing AI tools to identify areas of interest within a dynamic field of view is essential for objective behavioural evaluation of drivers. Video image classification and specifically image segmentation is a key technology to allow for the possibility of physiological and behavioural measurement of drivers. For example, to understand driver attention, one must measure where a driver is looking when driving and this requires segmentation of their field of view into relevant areas of interest, such as windows, mirrors, and dashboard. The present work addresses the challenge of dynamic field of view classification and shows the impact of transfer learning, a new AI tool, on segmentation accuracy. Results from this study demonstrate that transfer learning improves predictive performance by 0.02 to 0.20 Dice when large training sets were used. The resulting performance was >0.80 Dice for all classification tests of driver attention segmentation. This work showed that transfer learning also supported the use of smaller training sets while still providing adequate performance. This finding is key for applications where labeled training data is limited or costly to create. The present results expand the application space for deep learning-based image segmentation models.
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 | 0.001 | 0.002 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 0.001 |
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".