Diffusion-2BC: Hybrid Diffusion and Regression Training for Offline Behavior Cloning
Overview
Diffusion-2BC combines diffusion and behavior cloning for autonomous driving, reducing mask-distance error by 10% over diffusion baselines and 85% over standard cloning in controlled and CARLA environments, enhancing closed-loop reliability and action prediction.
Generated from attributed reports · Updated 47 minutes ago
Event evidence and corrections
0 attributed source owners. Ownership does not establish independent confirmation. Quantities are reported separately and are never added together.
No current evidence-backed claims. Missing information remains not reported.
Report timeline
Follow attributed reports and material updates.
- arXiv Robotics — research abstractsDiffusion-2BC: Hybrid Diffusion and Regression Training for Offline Behavior Cloning in Autonomous Driving
This research proposes Diffusion-2BC, which integrates a diffusion denoising objective with an auxiliary deterministic behavior-cloning loss over a shared visual encoder. Evaluated in controlled and CARLA environments, the method reduces mean mask-distance error by 10% compared to a diffusion baseline and 85% compared to standard behavior cloning, showing improved closed-loop reliability and multimodal action prediction.
Event coverage history
There is not enough continuous observation data to show a trend.
Timezone · UTC
Article dates follow your selected timezone. Briefing editions use Hong Kong time (UTC+8).