ASTAD: Asymmetric Style Transfer for Synthetic-to-Real Adaptation in Autonomous Driving
ASTAD:自动驾驶中合成到真实适应的非对称风格迁移
机构 * Department of Automation, Tsinghua University(自动化系,清华大学)
专题命中 感知 :autonomous driving(title,abstract);driving perception(abstract);分类 cs.CV
AI总结 提出ASTAD任务,针对合成数据有标注而真实数据无标注的不对称性,设计无训练两阶段框架ASTModel,通过粗语义先验提取和动态精炼实现类一致风格迁移,显著提升下游感知性能并加速推理。
Comments Accepted for publication at the 19th European Conference on Computer Vision (ECCV 2026)