TensorX
返回文献探索

Paper · arXiv 2311.13141

Diffusion360: Seamless 360 Degree Panoramic Image Generation based on Diffusion Models

Mengyang Feng, Jinlin Liu, Miaomiao Cui, Xuansong Xie

15 upvotesNovember 22, 2023arXiv 预印本
AI 摘要

A circular blending strategy is introduced to generate seamless 360-degree panoramic images using diffusion models for both text-to-360-panoramas and single-image-to-360-panoramas tasks.

diffusion models360-degree panoramic imagescircular blending strategydenoisingVAE decodingtext-to-360-panoramassingle-image-to-360-panoramas

Abstract

This is a technical report on the 360-degree panoramic image generation task based on diffusion models. Unlike ordinary 2D images, 360-degree panoramic images capture the entire 360^circtimes 180^circ field of view. So the rightmost and the leftmost sides of the 360 panoramic image should be continued, which is the main challenge in this field. However, the current diffusion pipeline is not appropriate for generating such a seamless 360-degree panoramic image. To this end, we propose a circular blending strategy on both the denoising and VAE decoding stages to maintain the geometry continuity. Based on this, we present two models for Text-to-360-panoramas and Single-Image-to-360-panoramas tasks. The code has been released as an open-source project at https://github.com/ArcherFMY/SD-T2I-360PanoImage{https://github.com/ArcherFMY/SD-T2I-360PanoImage} and https://www.modelscope.cn/models/damo/cv_diffusion_text-to-360panorama-image_generation/summary{ModelScope}

北京市昌平区探索星信息技术及软件开发工作室

京ICP备2026059466号