MVDiffusion: Enabling Holistic Multi-view Image Generation with Correspondence-Aware Diffusion

TL;DR

MVDiffusion is a novel diffusion-based method that generates consistent multi-view images from text prompts by incorporating correspondence-aware attention, effectively handling errors and enabling high-quality panorama and multi-view scene generation.

Contribution

It introduces a global, parallel multi-view image generation approach with correspondence-aware attention, improving consistency and quality over prior iterative warping methods.

Findings

High-resolution photorealistic panorama generation from limited training data

State-of-the-art multi-view scene texturing performance

Effective extrapolation to 360-degree views from single perspectives

Abstract

This paper introduces MVDiffusion, a simple yet effective method for generating consistent multi-view images from text prompts given pixel-to-pixel correspondences (e.g., perspective crops from a panorama or multi-view images given depth maps and poses). Unlike prior methods that rely on iterative image warping and inpainting, MVDiffusion simultaneously generates all images with a global awareness, effectively addressing the prevalent error accumulation issue. At its core, MVDiffusion processes perspective images in parallel with a pre-trained text-to-image diffusion model, while integrating novel correspondence-aware attention layers to facilitate cross-view interactions. For panorama generation, while only trained with 10k panoramas, MVDiffusion is able to generate high-resolution photorealistic images for arbitrary texts or extrapolate one perspective image to a 360-degree view. For multi-view depth-to-image generation, MVDiffusion demonstrates state-of-the-art performance for texturing a scene mesh.