DiffPano: Scalable and Consistent Text to Panorama Generation with Spherical Epipolar-Aware Diffusion

TL;DR

DiffPano introduces a scalable, consistent, and diverse text-driven panoramic image generation framework that leverages a large panoramic dataset and spherical epipolar-aware diffusion to improve multi-view consistency.

Contribution

The paper presents a novel large-scale panoramic dataset and a diffusion-based framework with spherical epipolar awareness for high-quality panoramic scene generation.

Findings

Successfully generates diverse panoramic images from unseen text and camera poses.

Ensures multi-view consistency through spherical epipolar-aware diffusion.

Achieves scalable and high-quality panoramic scene synthesis.

Abstract

Diffusion-based methods have achieved remarkable achievements in 2D image or 3D object generation, however, the generation of 3D scenes and even $360^{\circ}$ images remains constrained, due to the limited number of scene datasets, the complexity of 3D scenes themselves, and the difficulty of generating consistent multi-view images. To address these issues, we first establish a large-scale panoramic video-text dataset containing millions of consecutive panoramic keyframes with corresponding panoramic depths, camera poses, and text descriptions. Then, we propose a novel text-driven panoramic generation framework, termed DiffPano, to achieve scalable, consistent, and diverse panoramic scene generation. Specifically, benefiting from the powerful generative capabilities of stable diffusion, we fine-tune a single-view text-to-panorama diffusion model with LoRA on the established panoramic video-text dataset. We further design a spherical epipolar-aware multi-view diffusion model to ensure the multi-view consistency of the generated panoramic images. Extensive experiments demonstrate that DiffPano can generate scalable, consistent, and diverse panoramic images with given unseen text descriptions and camera poses.