Paper 2508.09983
Story2Board: A Training-Free Approach for Expressive Storyboard Generation
- Published
- Aug 2025
- Research lab
- Independent
- Citations
- 12
- GitHub
- 266 stars
01 In brief
Summary
Story2Board is a training-free framework for generating expressive storyboards from natural language.
It introduces two mechanisms: Latent Panel Anchoring (LPA), which pairs each panel with a shared reference prompt to maintain character consistency, and Reciprocal Attention Value Mixing (RAVM), which blends value vectors between semantically aligned tokens across panels.
The method uses an LLM to decompose stories into panel prompts and works with diffusion transformers like Flux and Stable Diffusion 3 without fine-tuning.
The authors introduce the Rich Storyboard Benchmark (100 open-domain stories) and a Scene Diversity metric to evaluate layout and pose variation.
Experiments show Story2Board achieves a better balance of prompt alignment, character consistency, and scene diversity than baselines like StoryDiffusion, IC-LoRA, and OminiControl.
A user study found it most preferred overall, though some baselines excelled in specific metrics.
Limitations include inheriting attention entanglement issues from base models.
02 From the paper
Abstract
We present Story2Board, a training-free framework for expressive storyboard generation from natural language. Existing methods narrowly focus on subject identity, overlooking key aspects of visual storytelling such as spatial composition, background evolution, and narrative pacing. To address this, we introduce a lightweight consistency framework composed of two components: Latent Panel Anchoring, which preserves a shared character reference across panels, and Reciprocal Attention Value Mixing, which softly blends visual features between token pairs with strong reciprocal attention. Together, these mechanisms enhance coherence without architectural changes or fine-tuning, enabling state-of-the-art diffusion models to generate visually diverse yet consistent storyboards. To structure generation, we use an off-the-shelf language model to convert free-form stories into grounded panel-level prompts. To evaluate, we propose the Rich Storyboard Benchmark, a suite of open-domain narratives designed to assess layout diversity and background-grounded storytelling, in addition to consistency. We also introduce a new Scene Diversity metric that quantifies spatial and pose variation across storyboards. Our qualitative and quantitative results, as well as a user study, show that Story2Board produces more dynamic, coherent, and narratively engaging storyboards than existing baselines.