Go to ICLR 2026 Conference homepageIPCP: Interpreter, Planner, Checker, and Painter Dialogue for Compositional Text-to-Image Generation
Keywords: Text-to-Image Generation, Multi Agent System
TL;DR: Multi-Agent system Compositional Text-to-Image Generation
Abstract: Text-to-image generation has advanced rapidly, but existing models still struggle with faithfully composing multiple objects and preserving their attributes in complex scenes. We propose IPCP, an interactive multi-agent dialogue framework with four specialized agents: Interpreter, Planner, Checker, and Painter that collaborate to improve compositional generation. The Interpreter adaptively decides between a direct text-to-image pathway and a layout-aware multi-agent process. In the layout-aware mode, it parses the prompt into attribute-rich object descriptors, ranks them by semantic salience, and groups objects with the same semantic priority level for joint generation.
Guided by the Interpreter, the Planner adopts a divide-and-conquer strategy, incrementally proposing layouts for objects with the same semantic priority level while grounding decisions in the evolving visual context of the canvas.
The Checker introduces an explicit error-correction mechanism by validating spatial consistency and attribute alignment, and refining the layouts before they are rendered.
Finally, the Painter synthesizes the image step by step, incorporating newly planned objects into the canvas to provide richer context for subsequent iterations. Together, these agents address three key challenges: reducing layout complexity, grounding planning in visual context, and enabling explicit error correction. Extensive experiments on compositional benchmarks GenEval and DPG-Bench demonstrate that IPCP substantially improves text–image alignment, spatial accuracy, and attribute binding compared to existing methods.
Primary Area: applications to computer vision, audio, language, and other modalities
Submission Number: 5241
Loading