This work reformulates the transition-level log-likelihood in Flow-GRPO into a token-wise representation natively aligned with DiT patch architectures, constructing spatially fine-grained importance sampling ratios and introduces a continuous spatial credit map derived from reward gradients.
Yun-Kai Yang, Yudong Zhang, Xin-Ying Chen et al.· 0 citations
UniEvo-RS, an omni-prompt unified RS segmentation framework equipped with representative exemplar-driven prototype evolution mechanism, unifies diverse prompting tasks, and enables training-free, progressive accuracy enhancement on unseen categories during batch annotation.
Kunquan Zhang, Peilang Li, Xikun Hu et al.· 0 citations
ControlRef is proposed, a highly efficient and precise multi-instance synthesis framework that achieves state-of-the-art visual fidelity and localization accuracy, while concurrently slashing inference latency by over 80% in sparse layouts and reducing memory overhead by 50% in dense scenarios.
Yun-Kai Yang, Yudong Zhang, Xin-Ying Chen et al.· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.