RECORD
Synthetic Data Generation Framework, Dataset, and Efficient Deep Model for Pedestrian Intention Prediction
Computer Vision and Pattern Recognition · Submitted 12 Jan 2024
abstract compiled · 2026-09-23 · no language-model review
The authors' main claim We will show how PedSynth complements widely used real-world datasets such as JAAD and PIE, so enabling more accurate models for C/NC prediction.
The compiled record
The paper's own sentences, arranged. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. Numbers in the tables are the authors', checked by code. No sentence in this record was written by a model.
Open the compiled record
The abstract, sentence by sentence
Pedestrian intention prediction is crucial for autonomous driving. In particular, knowing if pedestrians are going to cross in front of the ego-vehicle is core to performing safe and comfortable maneuvers. Creating accurate and fast models that predict such intentions from sequential images is challenging. A factor contributing to this is the lack of datasets with diverse crossing and non-crossing (C/NC) scenarios. We address this scarceness by introducing a framework, named ARCANE, which allows programmatically generating synthetic datasets consisting of C/NC video clip samples. As an example, we use ARCANE to generate a large and diverse dataset named PedSynth. We will show how PedSynth complements widely used real-world datasets such as JAAD and PIE, so enabling more accurate models for C/NC prediction. Considering the onboard deployment of C/NC prediction models, we also propose a deep model named PedGNN, which is fast and has a very low memory footprint. PedGNN is based on a GNN-GRU architecture that takes a sequence of pedestrian skeletons as input to predict crossing intentions.
background contribution method result interpretation main claim
| # | sentence | role | the authors' own work | states the main conclusion | strength |
|---|---|---|---|---|---|
| s0 | Pedestrian intention prediction is crucial for autonomous driving. | background 1.00 | 0.04 | 0.03 | asserted 1.00 |
| s1 | In particular, knowing if pedestrians are going to cross in front of the ego-vehicle is core to performing safe and comfortable maneuvers. | background 1.00 | 0.05 | 0.03 | asserted 1.00 |
| s2 | Creating accurate and fast models that predict such intentions from sequential images is challenging. | background 0.94 | 0.07 | 0.03 | asserted 1.00 |
| s3 | A factor contributing to this is the lack of datasets with diverse crossing and non-crossing (C/NC) scenarios. | gap 0.98 | 0.12 | 0.03 | asserted 0.89 |
| s4 | We address this scarceness by introducing a framework, named ARCANE, which allows programmatically generating synthetic datasets consisting of C/NC video clip samples. | contribution 0.88 | 0.94 | 0.12 | asserted 1.00 |
| s5 | As an example, we use ARCANE to generate a large and diverse dataset named PedSynth. | contribution 0.72 | 0.95 | 0.07 | asserted 1.00 |
| s6 | We will show how PedSynth complements widely used real-world datasets such as JAAD and PIE, so enabling more accurate models for C/NC prediction. | contribution 0.58 | 0.50 | 0.28 | asserted 0.91 |
| s7 | Considering the onboard deployment of C/NC prediction models, we also propose a deep model named PedGNN, which is fast and has a very low memory footprint. | contribution 0.81 | 0.91 | 0.11 | asserted 1.00 |
| s8 | PedGNN is based on a GNN-GRU architecture that takes a sequence of pedestrian skeletons as input to predict crossing intentions. | method 0.98 | 0.85 | 0.07 | asserted 1.00 |
What was examined, and what was not
- arXiv listing abstract v2
- full text
- figures and tables
- references and citing sentences
- relations to other papers
- P abstract
- S sections
- T body sentences
- C citations
- R cross-paper relations
The paper
Original abstract
Pedestrian intention prediction is crucial for autonomous driving. In particular, knowing if pedestrians are going to cross in front of the ego-vehicle is core to performing safe and comfortable maneuvers. Creating accurate and fast models that predict such intentions from sequential images is challenging. A factor contributing to this is the lack of datasets with diverse crossing and non-crossing (C/NC) scenarios. We address this scarceness by introducing a framework, named ARCANE, which allows programmatically generating synthetic datasets consisting of C/NC video clip samples. As an example, we use ARCANE to generate a large and diverse dataset named PedSynth. We will show how PedSynth complements widely used real-world datasets such as JAAD and PIE, so enabling more accurate models for C/NC prediction. Considering the onboard deployment of C/NC prediction models, we also propose a deep model named PedGNN, which is fast and has a very low memory footprint. PedGNN is based on a GNN-GRU architecture that takes a sequence of pedestrian skeletons as input to predict crossing intentions.
Human review
How the literature uses it
What this paper does with the arXiv papers it cites
What it cites · 3 sentences without a confident label
Cites as background 2 papers
-
Since then, there have been many works leveraging synthetic data to support the training of perception models or performing simulations [ref] ; being synth-to-real domain adaptation a core ingredient to encourage the use of synthetic data [ref] .
Synscapes: A Photorealistic Synthetic Dataset for Street Scene Parsing · Synthetic datasets focusing on C/NC prediction · 0.83
-
Since then, there have been many works leveraging synthetic data to support the training of perception models or performing simulations [ref] ; being synth-to-real domain adaptation a core ingredient to encourage the use of synthetic data [ref] .
Virtual KITTI 2 · Synthetic datasets focusing on C/NC prediction · 0.83
Discussion
Know something about this paper? Leave a comment, or highlight any passage above to comment on it.