PMMD: A pose-guided multi-view multi-modal diffusion for person generation

TL;DR

PMMD is a novel diffusion-based framework that synthesizes photorealistic human images with controllable pose and appearance by integrating multi-view references, pose maps, and text prompts, addressing occlusion and misalignment issues.

Contribution

The paper introduces PMMD, a pose-guided multi-view multimodal diffusion model with a multimodal encoder, ResCVA module, and cross modal fusion, improving person image synthesis quality.

Findings

Outperforms baselines in consistency and detail preservation

Effectively handles occlusions and pose misalignments

Demonstrates superior controllability in person image generation

Abstract

Generating consistent human images with controllable pose and appearance is essential for applications in virtual try on, image editing, and digital human creation. Current methods often suffer from occlusions, garment style drift, and pose misalignment. We propose Pose-guided Multi-view Multimodal Diffusion (PMMD), a diffusion framework that synthesizes photorealistic person images conditioned on multi-view references, pose maps, and text prompts. A multimodal encoder jointly models visual views, pose features, and semantic descriptions, which reduces cross modal discrepancy and improves identity fidelity. We further design a ResCVA module to enhance local detail while preserving global structure, and a cross modal fusion module that integrates image semantics with text throughout the denoising pipeline. Experiments on the DeepFashion MultiModal dataset show that PMMD outperforms representative baselines in consistency, detail preservation, and controllability. Project page and code are available at https://github.com/ZANMANGLOOPYE/PMMD.