An Explainable Attention Zone Estimation for Level 3 Autonomous Driving
Bibliographic record
Abstract
Accurately assessing the driver’s situational awareness is crucial in level 3 (<inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$L_{3}$ </tex-math></inline-formula>) autonomous driving, where the driver is in the loop. Estimating the attention zone provides essential information about the drivers’ on/off-road visual attention and determines their readiness to take over the control from the autonomous agent in complicated situations. This paper proposes a double-phase pipeline to improve the explainability and accuracy of the attention zone estimation using an intermediate gaze regression layer, where the true relationships between the input images and output zone labels are interpretable. The proposed GazeMobileNet, a lightweight deep neural network, in the first phase, achieved state-of-the-art performance in estimating the gaze vector in the MPIIGaze dataset, with MAE of 2.37 degrees. The model was used to extract the corresponding gaze vectors from the LISA V2, which is a driving dataset with the in-cabin attention zone labels. As LISA V2 does not contain gaze vector labels, an unsupervised clustering approach was proposed in the second phase to categorize the driver’s gaze vectors and map them to the corresponding attention zones. The proposed method demonstrated improved accuracy and robustness in the zone classification task. This model achieved the accuracies of 75.67% and 83.08% for attention zone estimation under “daytime without eyeglasses” and “nighttime without eyeglasses” capture conditions, respectively. Furthermore, the proposed model surpassed the recent research on that dataset by 73.11% and 74.02% accuracies under the “daytime with eyeglasses” and “nighttime with eyeglasses” capture conditions, respectively.
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 | 0.000 | 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 | 0.000 |
| 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".