Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

T2I-R1: Reinforcing Image Generation with Collaborative Semantic-level and Token-level CoT

About

Recent advancements in large language models have demonstrated how chain-of-thought (CoT) and reinforcement learning (RL) can improve performance. However, applying such reasoning strategies to the visual generation domain remains largely unexplored. In this paper, we present T2I-R1, a novel reasoning-enhanced text-to-image generation model, powered by RL with a bi-level CoT reasoning process. Specifically, we identify two levels of CoT that can be utilized to enhance different stages of generation: (1) the semantic-level CoT for high-level planning of the prompt and (2) the token-level CoT for low-level pixel processing during patch-by-patch generation. To better coordinate these two levels of CoT, we introduce BiCoT-GRPO with an ensemble of generation rewards, which seamlessly optimizes both generation CoTs within the same training step. By applying our reasoning strategies to the baseline model, Janus-Pro, we achieve superior performance with 13% improvement on T2I-CompBench and 19% improvement on the WISE benchmark, even surpassing the state-of-the-art model FLUX.1. Code is available at: https://github.com/CaraJ7/T2I-R1

Dongzhi Jiang, Ziyu Guo, Renrui Zhang, Zhuofan Zong, Hao Li, Le Zhuo, Shilin Yan, Pheng-Ann Heng, Hongsheng Li• 2025

Related benchmarks

TaskDatasetResultRank
Text-to-Image GenerationGenEval
Overall Score79
467
Text-to-Image GenerationT2I-CompBench
Shape Fidelity58.52
94
Text-to-Image GenerationGenEval
Overall Score79
68
Knowledge-grounded reasoningWISE
Overall Score54
45
Text-to-Image GenerationDPG-Bench (test)
Global Fidelity91.79
43
World Knowledge Image GenerationWISE
Overall Score54
39
Text-to-Image GenerationGenEval++
Color Accuracy68
35
Text-to-Image GenerationImagine-Bench
Attribute Shift Score5.85
33
Text-to-Image GenerationWISE (test)
Overall Score54
32
Text-to-Image GenerationOneIG EN
Alignment80.4
17
Showing 10 of 20 rows

Other info

Follow for update