Splatter Image: Ultra-Fast Single-View 3D Reconstruction

By Stanislaw Szymanowicz, Christian Rupprecht, Andrea Vedaldi. Computer Vision and Pattern Recognition, 2023

418 citations recorded by the source corpus.

We introduce the Splatter Image, an ultra-efficient approach for monocular 3D object reconstruction. Splatter Image is based on Gaussian Splatting, which allows fast and high-quality reconstruction of 3D scenes from multiple images. We apply Gaussian Splatting to monocular reconstruction by learning a neural network that, at test time, performs reconstruction in a feed-forward manner, at 38 FPS. Our main innovation is the surprisingly straightforward design of this network, which, using 2D operators, maps the input image to one 3D Gaussian per pixel. The resulting set of Gaussians thus has the form an image, the Splatter Image. We further extend the method take several images as input via cross-view attention. Owning to the speed of the renderer (588 FPS), we use a single GPU for training while generating entire images at each iteration to optimize perceptual metrics like LPIPS. On several synthetic, real, multi-category and large-scale benchmark datasets, we achieve better results in terms of PSNR, LPIPS, and other metries while training and evaluating much faster than prior works. Code, models and more results are available at https://szymanowiczs.github.io/splatter-image.

Subfields: Computer Graphics and Computer-Aided Design, Computer Vision and Pattern Recognition, Radiology, Nuclear Medicine and Imaging

View Paper: https://www.semanticscholar.org/paper/32032367b4e57d1fa355864982b62fb44b52cfc8

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Graphics and Computer-Aided Design

In Computer Vision and Pattern Recognition

In Radiology, Nuclear Medicine and Imaging

Citations in the corpus

LGM: Large Multi-View Gaussian Model for High-Resolution 3D Content Creation (2024, methodology): “We train an image-to-3D model with only one input views similar to splatter image [46], i.e. , without the multi-view generation step.”
A Survey on 3D Gaussian Splatting (2024, neutral): “These results indicate the effectiveness of introducing additional properties or structured information to model the deformation of Gaussians so as to model the scene dynamics.”
CAT3D: Create Anything in 3D with Multi-View Diffusion Models (2024, background): “Another line of research is to learn feed-forward models that take a few views as input, and output 3D representations directly, without an optimization process per instance [6, 55, 56, 57, 58, 59, 60].”
VGGT: Visual Geometry Grounded Transformer (2025, neutral): “Feed-forward Novel View Synthesis is progressing rapidly [4, 31, 34, 36, 70, 84, 95, 104].”
GRM: Large Gaussian Reconstruction Model for Efficient 3D Reconstruction and Generation (2024, methodology): “Leveraging various successful 2D GAN architectures [4, 23, 37, 39–42, 112], 3D GANs [5, 6, 22 , 24 , 65 , 67 , 77 , 82 , 87 , 88 , Some concurrent works, such as LGM [92], AGG [103], and Splatter Image [90], also use 3D Gaussians in a feed-forward model.”
GS-LRM: Large Reconstruction Model for 3D Gaussian Splatting (2024, background): “We found that the model has a certain ability to re-purposing points for hallucinat-ing unseen parts (also observed in [60]), but this capacity is limited and not guaranteed.”