From Prompting to Composing: A Spatial Canvas Interface for Poster Generation

2026-10-08

AI Summary

This study addresses the difficulty of precisely conveying 2D poster compositional intent through text prompts by proposing Compo, a novel system that introduces a spatial canvas interface to decouple intent specification from visual generation. This interface supports dual modes: explicit user construction and agent-driven automatic planning. Technically, Compo leverages pretrained image editing models, efficiently adapted through an automated data pipeline for fine-tuning. The system facilitates a paradigm shift from text prompting to spatial creation, significantly enhancing compositional controllability in poster generation while preserving high visual quality. Experimental results demonstrate that Compo outperforms existing general-purpose and task-specific baseline models.

Abstract

Text prompting is an indirect interface for poster generation, requiring users to encode inherently two-dimensional composition intent into a one-dimensional sequence of words. We introduce a Spatial Canvas Interface that enables users to directly compose generation intent in space through four complementary binding types: semantic, identity, text, and pixel, together with Text Specifications for individual elements and global appearance. Based on this interface, we develop Compo, a poster generation model adapted from a pretrained image editing model to understand Spatial Canvas inputs and Text Specifications. Compo supports both direct inference, where users explicitly construct the canvas, and agentic mode, where a high-level request is automatically translated into a planned Spatial Canvas. To train Compo, we develop a scalable pipeline that automatically constructs supervision data for different binding types and their combinations, enabling efficient adaptation without training a specialized poster generator from scratch. We further introduce a benchmark that evaluates adherence to individual binding types and their joint composition. Experiments show that Compo achieves stronger compositional controllability than both general-purpose image generation models and dedicated poster generation systems while maintaining high visual quality. By decoupling intent specification from visual generation, our work shifts poster generation from prompting toward composing.

Problem

• poster generation
• spatial composition
• text prompting
• compositional controllability
• user interface

Innovation

• Spatial Canvas Interface
• Poster Generation
• Compositional Controllability
• Agentic Mode
• Binding Types