Preprint
Aug 2026
Identity-Preserving Text-to-Video Generation via Agentic Enhancement and Semantic Repair
AESR introduces a global agentic prompt enhancement module, which learns model-specific prompting formats from official documentation, acquires human-centered video generation priors from human-interaction data, and accumulates test-domain identity-preserving generation experience into a reusable playbook through an agentic loop.
Jiayi Gao, Changcheng Hua, Jiaqi Tang et al.
· 0 citations