Pedestrian Crossing Intent Prediction Using Vision Transformers
Bibliographic record
Abstract
The prediction of pedestrian intentions is crucial and one of the most challenging problems for self-driving vehicles. For this reason, a fast, efficient, and robust vision-based model is required to predict pedestrian crossing as fast as possible and to prevent serious injuries or casualties that may occur. Transformers have rapidly replaced recurrent neural networks (RNN) based architectures for their better generalization and fast performance. Vision transformer (ViT) is a variant of transformers that has also proven to be efficient in image classification and has outperformed the state-of-the-art convolutional neural networks (CNN) when trained on large datasets. In this paper, a fully transformer-based architecture is presented to efficiently predict pedestrian intention with minimum latency. The proposed architecture is composed of two branches: the first branch handles the non-visual features while the second branch handles the visual features. The model is trained on the Joint Attention in Autonomous Driving (JAAD) dataset and different variants of the architecture are tested to find the optimal model. Experimental analysis shows that the proposed model outperforms all the previous state-of-the-art techniques, achieving the highest accuracy (83 %) and F1 score (64 %) on the testing dataset while maintaining the lowest processing time.
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 | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 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".