ZipSplat: Fewer Gaussians, Better Splats

• URL: http://arxiv.org/abs/2606.05102v1
• Date: Wed, 03 Jun 2026 17:04:30 GMT
• Title: ZipSplat: Fewer Gaussians, Better Splats
• Abstract summary: We propose ZipSplat, a token-based feed-forward model that decouples Gaussian placement from the pixel grid.<n>Cross- and self-attention refine these tokens, and a single trained model spans the quality-efficiency curve without retraining.<n>ZipSplat operates without ground-truth poses or intrinsics, yet sets a new state of the art on DL3DV and RealEstate10K with $sim6times$ fewer Gaussians than pixel-aligned methods.
• Score: 89.08529100444004
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Feed-forward 3D Gaussian Splatting methods reconstruct a scene from posed or pose-free images in a single forward pass, yet current approaches predict one Gaussian per input pixel, tying the representation budget to camera resolution rather than scene complexity. A flat wall and a richly textured object thus produce equally many Gaussians despite very different geometric needs. We propose ZipSplat, a token-based feed-forward model that decouples Gaussian placement from the pixel grid. A multi-view backbone extracts dense visual tokens, and k-means clustering compresses them into a compact set of scene tokens. Cross- and self-attention refine these tokens, and a lightweight MLP decodes each into a group of Gaussians with unconstrained 3D positions. Because clustering is applied at inference, a single trained model spans the quality-efficiency curve without retraining. ZipSplat operates without ground-truth poses or intrinsics, yet sets a new state of the art on DL3DV and RealEstate10K with ${\sim}6{\times}$ fewer Gaussians than pixel-aligned methods, surpassing the best pose-free baseline by 2.1dB and 1.2dB PSNR, respectively. It further generalizes zero-shot to Mip-NeRF360 and ScanNet++, outperforming all comparable baselines. Our project page is at ${\href{https://veichta.com/zipsplat}{https://veichta.com/zipsplat}}$.

Related papers

• InfiniSplat: Implicit Gaussian Decoding for Large-Baseline Monocular View Synthesis [49.87457834078904]
  InfiniSplat is a feed-forward single-image 3DGS framework that moves from a pixel-aligned representation toward a surface-aligned representation.<n>InfiniSplat achieves state-of-the-art performance compared with single-image feed-forward baselines.
  arXiv  Detail & Related papers  (2026-08-03T16:17:39Z)
• ATSplat: Compact Feed-forward 3D Gaussian Splatting with Adaptive Token Expansion [60.00320387969695]
  ATSplat is a feed-forward 3DGS framework that restores the adaptive allocation capability of 3DGS optimization.<n>From 12 input images at $512 times 960$ resolution, ATSplat completes reconstruction in less than a second using a single commercial GPU.
  arXiv  Detail & Related papers  (2026-07-22T17:54:44Z)
• AnchorSplat: Feed-Forward 3D Gaussian Splatting with 3D Geometric Priors [23.943522711585597]
  We propose AnchorSplat, a novel feed-forward 3DGS framework for scene-level reconstruction.<n> anchor-aligned Gaussian representation guided by 3D geometric priors.<n>Design substantially reduces the number of required Gaussians.
  arXiv  Detail & Related papers  (2026-04-08T13:04:54Z)
• YoNoSplat: You Only Need One Model for Feedforward 3D Gaussian Splatting [79.38712054342625]
  YoNoSplat is a feedforward model that reconstructs high-quality 3D Gaussian Splatting representations from an arbitrary number of images.<n>Our model is highly versatile, operating effectively with both posed and unposed, calibrated and uncalibrated inputs.<n>It achieves state-of-the-art performance on standard benchmarks in both pose-free and pose-dependent settings.
  arXiv  Detail & Related papers  (2025-11-10T17:21:54Z)
• VolSplat: Rethinking Feed-Forward 3D Gaussian Splatting with Voxel-Aligned Prediction [45.95623374754385]
  VolSplat is a new multi-view feed-forward paradigm that replaces pixel alignment with voxel-aligned Gaussians.<n>It overcomes pixel alignment's reliance on error-prone 2D feature matching, ensuring robust multi-view consistency.<n> Experiments on widely used benchmarks including RealEstate10K and ScanNet demonstrate that VolSplat achieves state-of-the-art performance.
  arXiv  Detail & Related papers  (2025-09-23T17:59:02Z)
• No Pose at All: Self-Supervised Pose-Free 3D Gaussian Splatting from Sparse Views [17.221166075016257]
  SPFSplat is an efficient framework for 3D Gaussian splatting from sparse multi-view images.<n>It employs a shared feature extraction backbone, enabling simultaneous prediction of 3D Gaussian primitives and camera poses.<n>It achieves state-of-the-art performance in novel view synthesis even under significant viewpoint changes and limited image overlap.
  arXiv  Detail & Related papers  (2025-08-02T03:19:13Z)
• AnySplat: Feed-forward 3D Gaussian Splatting from Unconstrained Views [68.94737256959661]
  AnySplat is a feed forward network for novel view synthesis from uncalibrated image collections.<n>A single forward pass yields a set of 3D Gaussian primitives encoding both scene geometry and appearance.<n>In extensive zero shot evaluations, AnySplat matches the quality of pose aware baselines in both sparse and dense view scenarios.
  arXiv  Detail & Related papers  (2025-05-29T17:49:56Z)
• LeanGaussian: Breaking Pixel or Point Cloud Correspondence in Modeling 3D Gaussians [11.71048049090424]
  We introduce LeanGaussian, a novel approach that treats each query in deformable Transformer as one 3D Gaussian ellipsoid.<n>We leverage deformable decoder to iteratively refine the Gaussians layer-by-layer with the image features as keys and values.<n>Our approach outperforms prior methods by approximately 6.1%, achieving a PSNR of 25.44 and 22.36, respectively.
  arXiv  Detail & Related papers  (2024-04-25T04:18:59Z)
• MVSplat: Efficient 3D Gaussian Splatting from Sparse Multi-View Images [102.7646120414055]
  We introduce MVSplat, an efficient model that, given sparse multi-view images as input, predicts clean feed-forward 3D Gaussians. On the large-scale RealEstate10K and ACID benchmarks, MVSplat achieves state-of-the-art performance with the fastest feed-forward inference speed (22fps)
  arXiv  Detail & Related papers  (2024-03-21T17:59:58Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.