Research map / Large language models / Agents, code and math
OD-RASE: Ontology-Driven Risk Assessment and Safety Enhancement for Autonomous Driving
Kota Shimomura, Masaki Nambata, Atsuya Ishikawa, Ryota Mimura, Koki Inoue, Takayoshi Yamashita, Takayuki Kawabuchi
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add ontology-driven data filtering of GPT-4o labels (ablation) | OD-RASE · Mapillary (improvement proposals) | F1: 44.26 -> 70.26 | Improved |
| add ontology-driven data filtering of GPT-4o labels (ablation) | OD-RASE · Mapillary (improvement proposals) | accuracy: 0.00 -> 42.14 | Improved |
| add ontology-driven data filtering of GPT-4o labels (ablation) | OD-RASE · Mapillary (road structures) | F1: 52.63 -> 79.99 | Improved |
| add text input to image-only model (ablation) | OD-RASE · Mapillary (improvement proposals) | accuracy: 34.50 -> 42.14 | Improved |
| add image input to text-only model (ablation) | OD-RASE · Mapillary (improvement proposals) | accuracy: 40.63 -> 42.14 | Improved |
| add image input to text-only model (ablation) | OD-RASE · Mapillary (improvement proposals) | F1: 68.50 -> 70.26 | Improved |
| add the grounding block (cross-attention) over plain CLIP (ablation) | OD-RASE · Mapillary (road structures) | F1: 62.55 -> 79.99 | Improved |
| add the grounding block over plain Long-CLIP (ablation) | OD-RASE · Mapillary (road structures) | accuracy: 28.22 -> 65.09 | Improved |
| use Flan-T5-xl instead of RoBERTa-Base as text encoder (ablation) | OD-RASE · Mapillary (improvement proposals) | accuracy: 42.14 -> 0.00 | Hurt |
| use Long-CLIP vision encoder instead of ResNet-50 (ablation) | OD-RASE · Mapillary (improvement proposals) | F1: 64.98 -> 70.26 | Improved |
| use OD-RASE instead of GPT-4o zero-shot (BDD100K to Mapillary) | OD-RASE · Mapillary test (zero-shot) | F1: 34.27 -> 68.32 | Improved |
| use OD-RASE instead of GPT-4o zero-shot (Mapillary to BDD100K) | OD-RASE · BDD100K test (zero-shot) | F1: 29.38 -> 76.00 | Improved |
Closest papers by meaning
- SafeAuto: Knowledge-Enhanced Safe Autonomous Driving with Multimodal Foundation Models (ICML 2025 · similarity 0.85)
- Dual-View Interaction-Aware Lane Change Prediction for Autonomous Driving (AAAI 2025 · similarity 0.84)
- Hybrid-Driving: An Autonomous Driving Decision Framework Integrating Large Language Models, Knowledge Graphs and Driving Rules (AAAI 2025 · similarity 0.83)
- RoadSceneBench: A Lightweight Benchmark for Mid-Level Road Scene Understanding (CVPR 2026 · similarity 0.83)
- AIDE: An Automatic Data Engine for Object Detection in Autonomous Driving (CVPR 2024 · similarity 0.83)
- Driving by the Rules: A Benchmark for Integrating Traffic Sign Regulations into Vectorized HD Map (CVPR 2025 · similarity 0.83)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
