Research map / Detection and segmentation / Open-set and few-shot detection
Advisable Learning for Self-Driving Vehicles by Internalizing Observation-to-Action Rules
Jinkyu Kim, Suhong Moon, Anna Rohrbach, Trevor Darrell, John Canny
Check 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 long-term (global) observation-action advice to discrete-command goal-conditioned model | goal-conditioned driving model D · BDD-X control prediction | ADE median with speed inputs: 1.35 -> 0.81 | Improved |
| replace ConvNet from scratch with object-centric visual encoder (semantic segmentation + RoI pooling) | model F · BDD-X control prediction | ADE median with speed inputs: 0.81 -> 0.65 | Improved |
| object-centric visual encoder | model F · BDD-X control prediction | ADE median without speed inputs: 2.14 -> 1.93 | Improved |
| add long-term advice to discrete-command model | model D · BDD-X control prediction | ADE median without speed inputs: 2.28 -> 2.14 | Mixed |
| object-centric visual encoder with semantic segmentation and RoI pooling, observation-to-action module | observation generation and action generation · BDD-X textual observation generation | CIDEr-D: 70.1 -> 98.2 (vs best baseline) | Improved |
| full advisable model with observation-to-action module | textual action command generation · BDD-X textual action command generation | CIDEr-D: 215.8 -> 338.5 | Improved |
| learned sim-based advice (red light, stop sign) given at test time in Carla | driving model in Carla simulator · Carla stopping scenarios | failure rate | Improved |
| add visual and textual explanations to the driving system | user trust study · Carla human study | user trust Likert (1-5): slight improvement | Improved |
Closest papers by meaning
- Grounding Human-To-Vehicle Advice for Self-Driving Vehicles (CVPR 2019 · similarity 0.91)
- Learning by Watching (CVPR 2021 · similarity 0.90)
- Continuously Learning, Adapting, and Improving: A Dual-Process Approach to Autonomous Driving (NeurIPS 2024 · similarity 0.89)
- Prioritizing Perception-Guided Self-Supervision: A New Paradigm for Causal Modeling in End-to-End Autonomous Driving (NeurIPS 2025 · similarity 0.89)
- Less Language, More Latents: Annotation-Efficient VLAs for Driving (NeurIPS 2026 · similarity 0.88)
- SafeAuto: Knowledge-Enhanced Safe Autonomous Driving with Multimodal Foundation Models (ICML 2025 · similarity 0.87)
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.
