# ConsistCompose: Unified Multimodal Layout Control for Image Composition

Source: [ConsistCompose: Unified Multimodal Layout Control for Image Composition](<https://www.sota2.com/research/paper/69253808d6b0dd334292f3c0>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-03-18T05:00:11.457Z

Authors: Xuanke Shi, Boxuan Li, Xiaoyang Han, Zhongang Cai, Lei Yang, Quan Wang, Dahua Lin

Published: 2025-11-23T08:14:53.000Z

## Abstract

Unified multimodal models that couple visual understanding with image generation have advanced rapidly, yet most systems still focus on visual grounding-aligning language with image regions-while their generative counterpart, linguistic-embedded layout-grounded generation (LELG) for layout-controllable multi-instance generation, remains underexplored and limits precise compositional control. We present ConsistCompose, a unified multimodal framework that embeds layout coordinates directly into language prompts, enabling layout-controlled multi-instance image generation from Interleaved Image-Text within a single generative interface. We further construct ConsistCompose3M, a 3.4M multi-instance generation dataset with layout and identity annotations (2.6M text-guided and 0.8M image-guided data pairs) that provides large-scale supervision for layout-conditioned generation. Within this framework, LELG is instantiated through instance-coordinate binding prompts and coordinate-aware classifier-free guidance, which translate linguistic layout cues into precise spatial control without task-specific branches. Experiments on COCO-Position and MS-Bench show that ConsistCompose substantially improves spatial accuracy over layout-controlled baselines while preserving identity fidelity and competitive general multimodal understanding, establishing a unified paradigm for layout-controllable multimodal image generation.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2511.18333>)

[PDF](<https://arxiv.org/pdf/2511.18333.pdf>)