TAGRPO: Boosting GRPO on Image-to-Video Generation with Direct Trajectory Alignment 文章

ArXiv CS.CV2026-05-27NEWSen作者: Jin Wang, Jianxiang Lu, Guangzheng Xu, Comi Chen, Haoyu Yang, Linqing Wang, Peng Chen, Mingtao Chen, Zhichao Hu, Longhuang Wu, Shuai Shao, Qinglin Lu, Ping Luo
查看原文 →
关系图谱

TAGRPO: Boosting GRPO on Image-to-Video Generation with Direct Trajectory Alignment · 相关技术

相关技术

contrastive learningginODEPLA递归自我改进后训练divide-and-conquer partitioningalignmentText-to-imageStraight-Through EstimatorStanReferring expression comprehension (REC)RPAParts-of-Speech (POS) tagsPPRMITHISGroupGroup Relative Policy OptimizationGRPOForFFIANNAGI