SCOPE: Semantic Bounded Credit Optimization for Prompt Engineering in Text-to-image Generation
초록
Reinforcement learning (RL) enables language models to improve text-to-image (T2I) generation by rewriting prompts for a frozen image generator. Existing methods typically apply a single signal derived from the image to individual subword tokens, although objects, counts, attributes, and relations form coherent semantic spans. We propose SCOPE, an objective based on Group Relative Policy Optimization (GRPO) that aligns update weighting with this structure. The rewriter jointly generates an enhanced prompt and its semantic partition. SCOPE pools shifts in policy log probabilities within each span to derive a shared coefficient for its token losses while retaining a single advantage for each completion. It requires neither separate span rewards nor additional image renders. Across Qwen-Image, SD3, and FLUX.1-dev, SCOPE achieves average relative gains over vanilla GRPO of 9.48% on GeneVal and 7.62% on T2I-CompBench. Ablation studies show that both semantic boundaries and pooling within spans contribute to these gains, supporting semantic structure as a practical inductive bias for prompt engineering with RL.
저자 (8명)
- Yufei Jiao — LinkedIn 검색
- Jinfa Huang — LinkedIn 검색
- Zeyue Xue — LinkedIn 검색
- Tianyu Xie — LinkedIn 검색
- Yuexiao Ma — LinkedIn 검색
- Rongrong Ji — LinkedIn 검색
- Xiawu Zheng — LinkedIn 검색
- Jiebo Luo — LinkedIn 검색
저자 LinkedIn 변경 추적은 추후 자동화 예정입니다. 현재는 검색 링크를 제공합니다.