论文概要
研究领域: CV 作者: Haoran Feng, Ruiyang Zhang, Longyi Zhang, Dizhe Zhang, Lu Qi 发布时间: 2026-07-09 arXiv: 2607.08765
中文摘要
本工作提出Canvas360，一种用于上下文全景生成的两阶段框架，将几何感知预训练与下游任务特定微调相结合。为解决上下文全景任务缺乏大规模高质量训练数据的问题，我们提出Canvas360Dataset，包含100万个高质量配对全景样本，用于风格迁移、修复、外扩和编辑，从而实现跨多样化上下文生成场景的有效监督。在建模方面，Canvas360通过并行深度生成、速度循环填充和相似性损失正则化增强文本到全景生成，使模型能够学习几何感知表征、捕获物体畸变细节并提高几何一致性和全局连贯性。此外，凭借强大的全景先验，Canvas360实现了统一的上下文全景生成框架，通过token级拼接支持多样化下游任务，在任务覆盖范围和建模灵活性方面超越了先前方法。大量实验表明，Canvas360提升了全景图像保真度，在全景特定的FAED指标上表现尤为突出，在所有报告的定量评估中取得了有竞争力或领先的结果。更多信息请访问项目页面：https://zry000.github.io/Canvas360/
原文摘要
In this work, we present Canvas360, a two-stage framework for in-context panoramic generation that combines geometry-aware pretraining with downstream task-specific fine-tuning. To address the lack of large-scale, high-quality training data tailored to in-context panoramic tasks, we propose Canvas360Dataset, a collection of 1M high-quality paired panoramic samples for style transfer, inpainting, outpainting, and editing, enabling effective supervision across diverse in-context generation scenarios. On the modeling side, Canvas360 enhances text-to-panorama generation through parallel depth generation, velocity circular padding, and similarity loss regularization, enabling the model to learn geometry-aware representations, capture object distortion details, and improve geometric consistency …
— 自动采集于 2026-07-12
#论文 #arXiv #CV #小凯