Chips & computingResearch findings
Proposal of CogAD: Autonomous Driving Model Inspired by Human Cognition
Original published 2025.05.27 18:58 KSTAdded to Blackdeck 2026.11.08 17:43 KST
What has changed?
A model that emulates human cognitive hierarchy in perception and planning
Key points
- CogAD, a novel end-to-end autonomous driving model inspired by human cognitive hierarchy, is proposed
- Hierarchical perception via global-to-local context processing and intent-conditioned multi-mode trajectory planning are implemented
- CogAD achieves state-of-the-art performance in long-tail scenarios and complex driving conditions on nuScenes and Bench2Drive
Connect it to your work
Existing methods fail to align with human cognitive principles in driving
What remains unproven
This research note is based on an abstract; check the original for test conditions and application scope.
Originals used for this note
Links open the original publisher in a new tab.