TY  - RPRT
TI  - FinPercep-RM: A Fine-grained Reward Model and Co-evolutionary Curriculum for RL-based Real-world Super-Resolution
AU  - Yidi Liu
AU  - Zihao Fan
AU  - Jie Huang
AU  - Jie Xiao
AU  - Dong Li
AU  - Wenlong Zhang
AU  - Lei Bai
AU  - Xueyang Fu
AU  - Zheng-Jun Zha
PY  - 2026
UR  - https://arxiv.org/abs/2512.22647
ID  - 2512.22647
ER  -