Off The Grid: Detection of Primitives for Feed-Forward 3D Gaussian Splatting

Arthur Moreau, Richard Shaw, Michal Nazarczuk, Jisu Shin, Thomas Tanay,
Zhensong Zhang, Songcen Xu, Eduardo Pérez-Pellitero

Huawei Noah's Ark Lab
Interpolate start reference image.

Off The Grid is a new method for Feed-Forward (Pose-Free) 3D Gaussian Splatting. Instead of placing primitives in a regular grid, we detect subpixel Gaussians locations. Our model produce more accurate 3D Gaussians models while using much less primitives.

Abstract

Feed-forward 3D Gaussian Splatting (3DGS) models enable real-time scene generation but are hindered by suboptimal pixel-aligned primitive placement, which relies on a dense, rigid grid and limits both quality and efficiency. We introduce a new feed-forward architecture that detects 3D Gaussian primitives at a sub-pixel level, replacing the pixel grid with an adaptive, "Off The Grid" distribution. Inspired by keypoint detection, our multi-resolution decoder learns to distribute primitives across image patches. This module is trained end-to-end with a 3D reconstruction backbone using self-supervised learning. Our resulting pose-free model generates photorealistic scenes in seconds, achieving state-of-the-art novel view synthesis for feed-forward models. It outperforms competitors while using far fewer primitives, demonstrating a more accurate and efficient allocation that captures fine details and reduces artifacts. Moreover, we observe that by learning to render 3D Gaussians, our 3D reconstruction backbone improves camera pose estimation, suggesting opportunities to train these foundational models without labels.

Video

Novel View Synthesis

Models created using 6 images from DL3DV-Benchmark scenes. We compare interpolated and extrapolated view synthesis between AnySplat and our model. Off The Grid presents higher quality geometry and less artifacts than AnySlat, especially visible during the spiral motion.

Method

We fine-tune VGGT to predict and render 3D Gaussians in a self-supervised loop that does not require 3D annotations for training.

OffTheGrid pipeline

Our 3D Gaussians Decoder learns to detect and describe 3D primitives from 2D images, depths and latent features.

OffTheGrid pipeline

Effect of Fine-Tuning on 3D geometry

Fine-tuning VGGT for rendering with our method enables to avoid depth estimation failure cases in specular or sky areas.

OffTheGrid pipeline

We also observe that our fine-tuning improves camera pose and focal length estimations, check the paper for more details !

BibTeX

@article{moreau2023human,
      title={Off The Grid: Detection of Primitives for Feed-Forward 3D Gaussian Splatting},
      author={Moreau, Arthur and Shaw, Richard and Nazarczuk, Michal  and Shin, Jisu and Tanay, Thomas and Zhang, Zhensong, and Xu, Songcen and P{\'e}rez-Pellitero, Eduardo},
      journal={arXiv preprint},
      year={2025}
    }