PanSplat: 4K Panorama Synthesis with Feed-Forward Gaussian Splatting

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Zhang, Cheng, Xu, Haofei, Wu, Qianyi, Gambardella, Camilo Cruz, Phung, Dinh, Cai, Jianfei
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908279019405312
author Zhang, Cheng
Xu, Haofei
Wu, Qianyi
Gambardella, Camilo Cruz
Phung, Dinh
Cai, Jianfei
author_facet Zhang, Cheng
Xu, Haofei
Wu, Qianyi
Gambardella, Camilo Cruz
Phung, Dinh
Cai, Jianfei
contents With the advent of portable 360° cameras, panorama has gained significant attention in applications like virtual reality (VR), virtual tours, robotics, and autonomous driving. As a result, wide-baseline panorama view synthesis has emerged as a vital task, where high resolution, fast inference, and memory efficiency are essential. Nevertheless, existing methods are typically constrained to lower resolutions (512 $\times$ 1024) due to demanding memory and computational requirements. In this paper, we present PanSplat, a generalizable, feed-forward approach that efficiently supports resolution up to 4K (2048 $\times$ 4096). Our approach features a tailored spherical 3D Gaussian pyramid with a Fibonacci lattice arrangement, enhancing image quality while reducing information redundancy. To accommodate the demands of high resolution, we propose a pipeline that integrates a hierarchical spherical cost volume and Gaussian heads with local operations, enabling two-step deferred backpropagation for memory-efficient training on a single A100 GPU. Experiments demonstrate that PanSplat achieves state-of-the-art results with superior efficiency and image quality across both synthetic and real-world datasets. Code is available at https://github.com/chengzhag/PanSplat.
format Preprint
id arxiv_https___arxiv_org_abs_2412_12096
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle PanSplat: 4K Panorama Synthesis with Feed-Forward Gaussian Splatting
Zhang, Cheng
Xu, Haofei
Wu, Qianyi
Gambardella, Camilo Cruz
Phung, Dinh
Cai, Jianfei
Computer Vision and Pattern Recognition
With the advent of portable 360° cameras, panorama has gained significant attention in applications like virtual reality (VR), virtual tours, robotics, and autonomous driving. As a result, wide-baseline panorama view synthesis has emerged as a vital task, where high resolution, fast inference, and memory efficiency are essential. Nevertheless, existing methods are typically constrained to lower resolutions (512 $\times$ 1024) due to demanding memory and computational requirements. In this paper, we present PanSplat, a generalizable, feed-forward approach that efficiently supports resolution up to 4K (2048 $\times$ 4096). Our approach features a tailored spherical 3D Gaussian pyramid with a Fibonacci lattice arrangement, enhancing image quality while reducing information redundancy. To accommodate the demands of high resolution, we propose a pipeline that integrates a hierarchical spherical cost volume and Gaussian heads with local operations, enabling two-step deferred backpropagation for memory-efficient training on a single A100 GPU. Experiments demonstrate that PanSplat achieves state-of-the-art results with superior efficiency and image quality across both synthetic and real-world datasets. Code is available at https://github.com/chengzhag/PanSplat.
title PanSplat: 4K Panorama Synthesis with Feed-Forward Gaussian Splatting
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2412.12096