CoF-T2I: Video Models as Pure Visual Reasoners for Text-to-Image Generation

TL;DR

This paper introduces CoF-T2I, a novel model that leverages video reasoning techniques for improved text-to-image generation by using explicit intermediate reasoning steps and a new dataset of visual trajectories.

Contribution

The paper presents CoF-T2I, integrating Chain-of-Frame reasoning into T2I generation with progressive refinement and a new dataset, enhancing interpretability and quality.

Findings

Outperforms base video models in T2I tasks

Achieves 0.86 on GenEval benchmark

Reaches 7.468 on Imagine-Bench benchmark

Abstract

Recent video generation models have revealed the emergence of Chain-of-Frame (CoF) reasoning, enabling frame-by-frame visual inference. With this capability, video models have been successfully applied to various visual tasks (e.g., maze solving, visual puzzles). However, their potential to enhance text-to-image (T2I) generation remains largely unexplored due to the absence of a clearly defined visual reasoning starting point and interpretable intermediate states in the T2I generation process. To bridge this gap, we propose CoF-T2I, a model that integrates CoF reasoning into T2I generation via progressive visual refinement, where intermediate frames act as explicit reasoning steps and the final frame is taken as output. To establish such an explicit generation process, we curate CoF-Evol-Instruct, a dataset of CoF trajectories that model the generation process from semantics to aesthetics. To further improve quality and avoid motion artifacts, we enable independent encoding operation for each frame. Experiments show that CoF-T2I significantly outperforms the base video model and achieves competitive performance on challenging benchmarks, reaching 0.86 on GenEval and 7.468 on Imagine-Bench. These results indicate the substantial promise of video models for advancing high-quality text-to-image generation.