Enhancing In-context Panoramic Generation via Geometric-aware Pretraining

Hugging Face Daily Papers Papers
panoramic-generation geometric-aware-pretraining in-context-generation depth-generation dataset-creation

Summary

Canvas360 is a two-stage framework for in-context panoramic generation that combines geometry-aware pretraining with fine-tuning, featuring a large-scale dataset and novel modeling techniques for improved geometric consistency and global coherence.

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 and global coherence. Furthermore, empowered by strong panoramic priors, Canvas360 enables a unified in-context panoramic generation framework that supports diverse downstream tasks via token-level concatenation, surpassing prior methods in both task coverage and modeling flexibility. Extensive experiments show that Canvas360 improves panoramic image fidelity, achieving particularly strong performance on the panorama-specific FAED metric and competitive or leading results across the reported quantitative evaluations. More information can be found on our project page: https://zry000.github.io/Canvas360/
Original Article
View Cached Full Text

Cached at: 07/10/26, 10:10 AM

Paper page - Enhancing In-context Panoramic Generation via Geometric-aware Pretraining

Source: https://huggingface.co/papers/2607.08765

Abstract

Canvas360 is a two-stage framework for in-context panoramic generation that combines geometry-aware pretraining with fine-tuning, featuring a large-scale dataset and novel modeling techniques for improved geometric consistency and global coherence.

In this work, we present Canvas360, a two-stage framework forin-context panoramic generationthat combinesgeometry-aware pretrainingwithdownstream 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 throughparallel depth generation,velocity circular padding, andsimilarity loss regularization, enabling the model to learn geometry-aware representations, capture object distortion details, and improvegeometric consistencyandglobal coherence. Furthermore, empowered by strong panoramic priors, Canvas360 enables a unifiedin-context panoramic generationframework that supports diverse downstream tasks viatoken-level concatenation, surpassing prior methods in both task coverage and modeling flexibility. Extensive experiments show that Canvas360 improvespanoramic image fidelity, achieving particularly strong performance on thepanorama-specific FAED metricand competitive or leading results across the reported quantitative evaluations. More information can be found on our project page: https://zry000.github.io/Canvas360/

View arXiv pageView PDFProject pageGitHub5Add to collection

Get this paper in your agent:

hf papers read 2607\.08765

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2607.08765 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2607.08765 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2607.08765 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles

PanoWorld: Real-World Panoramic Generation

Hugging Face Daily Papers

PanoWorld proposes a method for long-range memory in panoramic world models using rotation-equivariant representations, with a three-stage training pipeline and a new large-scale dataset World360. The model outperforms alternatives by a large margin.

Unified Panoramic Geometry Estimation via Multi-View Foundation Models

Hugging Face Daily Papers

PaGeR adapts the multi-view perspective foundation model Depth Anything 3 to predict scale-invariant and metric depth, surface normals, and sky segmentation from a single equirectangular image, using a fixed cubemap representation that keeps VRAM and runtime constant. The paper also releases the ZüriPano and PanoInfinigen datasets.

PanoWorld: Towards Spatial Supersensing in 360^circ Panorama World

Hugging Face Daily Papers

PanoWorld introduces spherical spatial cross-attention for panoramic reasoning, addressing limitations of MLLMs in 360-degree spatial understanding. It builds a large-scale pipeline for geometry-aware supervision and proposes a diagnostic benchmark, achieving state-of-the-art results on multiple benchmarks.