{"ID":5551721,"CreatedAt":"2026-07-02T01:54:51.863792489Z","UpdatedAt":"2026-07-04T11:43:16.960862486Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2607.00832","arxiv_id":"2607.00832","title":"Pano2World: End-to-End 3D Generation via Unified Multi-View Sequences","abstract":"A single panorama captures the full visual sphere from one camera center, yet confines users to looking around in place without enabling true scene exploration. Converting a single panorama into a persistent, renderable 3D representation for free-viewpoint navigation has attracted growing interest; existing methods either adopt iterative per-view completion that propagates inpainting results to update the underlying geometry, leading to progressive error accumulation and cumbersome multi-step pipelines, or leverage the temporal consistency priors of video generation models, yet the continuous-trajectory constraint intrinsic to such models limits their flexibility in covering scenes from multiple directions simultaneously. We present Pano2World, which takes a single indoor panorama as input and directly outputs a persistent, explorable 3D Gaussian scene. Given the source panorama, Pano2World first reconstructs a coarse 3D Gaussian proxy and renders it at adaptively sampled nearby poses to obtain geometrically aligned guidance panoramas; a panoramic diffusion model then jointly denoises all target views via View-Aware Attention Routing, where each target view simultaneously receives geometric constraints from its corresponding guidance panorama and global semantic guidance from the source panorama, naturally enforcing cross-view consistency. To avoid the information loss incurred by decoding the multi-view hidden features formed during joint denoising back to the pixel domain via VAE, we introduce Latent Feature Adapter, a geometry-aware bridge module that directly distills these hidden features into a scene latent, subsequently decoded into the final 3D Gaussian scene. Experiments demonstrate that Pano2World significantly outperforms existing methods on the multi-position panoramic novel-view synthesis benchmark.","short_abstract":"A single panorama captures the full visual sphere from one camera center, yet confines users to looking around in place without enabling true scene exploration. Converting a single panorama into a persistent, renderable 3D representation for free-viewpoint navigation has attracted growing interest; existing methods eit...","url_abs":"https://arxiv.org/abs/2607.00832","url_pdf":"https://arxiv.org/pdf/2607.00832v1","authors":"[\"Zhenjia Li\",\"Jinrang Jia\",\"Yifeng Shi\"]","published":"2026-07-01T11:54:02Z","proceeding":"cs.CV","tasks":"[\"cs.CV\",\"cs.AI\"]","methods":"[\"Diffusion Model\",\"LoRA\",\"Variational Autoencoder\"]","has_code":false}