Free-Range Gaussians: Non-Grid-Aligned Generative 3D Gaussian Reconstruction
Authors: Ahan Shabanov, Peter Hedman, Ethan Weber, Zhengqin Li, Denis Rozumny, Gael Le Lan, Naina Dhingra, Lei Luo, Andrea Vedaldi, Christian Richardt, Andrea Tagliasacchi, Bo Zhu, Numair Khan
Organizations: Meta Reality Labs · Simon Fraser University · University of Oxford · University of Toronto
We present Free-Range Gaussians, a multi-view reconstruction method that predicts non-pixel, non-voxel-aligned 3D Gaussians from as few as four images. This is done through flow matching over Gaussian parameters. Our generative formulation of reconstruction allows the model to be supervised with non-grid-aligned 3D data, and enables it to synthesize plausible content in unobserved regions. Thus, it improves on prior methods that produce highly redundant grid-aligned Gaussians, and suffer from holes or blurry conditional means in unobserved regions. To handle the number of Gaussians needed for high-quality results, we introduce a hierarchical patching scheme to group spatially related Gaussians into joint transformer tokens, halving the sequence length while preserving structure. We further propose a timestep-weighted rendering loss during training, and photometric gradient guidance and classifier-free guidance at inference to improve fidelity. Experiments on Objaverse and Google Scanned Objects show consistent improvements over pixel and voxel-aligned methods while using significantly fewer Gaussians, with large gains when input views leave parts of the object unobserved.
Figures & tables
Figure 1: Non-grid-aligned generative reconstruction. From four sparse views, our method produces a complete 3D Gaussian reconstruction via flow matching, achieving state-of-the-art quality for both observed (high PSNR) and unobserved regions (low FID) with ∼10× fewer Gaussians than grid-aligned methods.
Figure 2: Generation vs. Reconstruction. State-of-the-art generative methods like TRELLIS [ 60 ] produce sharp output but suffer from weak input conditioning as they fail to preserve object orientation, and hallucinate geometry and appearance even in observed regions. Our results remain aligned with input-view poses, geometry, and appearance. Please see the supplementary material for additional results.
Figure 3: Method overview. We pose reconstruction as a generative process based on flow matching. Conditioned on sparse image features, a diffusion transformer denoises a set of non-grid-aligned 3D Gaussians from random noise ( Section 3.1 ). A level-of-detail tree enables Gaussian patchification for tractable sequence lengths ( Section 3.2 ). A timestep-weighted rendering loss, photometric gradient guidance, and classifier-free guidance steer the output toward consistent reconstructions ( Section 3.3 ).
Figure 4: Generative reconstruction is efficient and complete. We visualize reconstruction density by rendering scaled-down Gaussians with random colors. Pixel-aligned methods like GS-LRM [ 69 ] leave holes in unobserved regions despite a highly redundant representation. Voxel-aligned methods like LaRa [ 4 ] are more complete but predict blurry conditional means for unseen parts. Our generative approach produces sharp, plausible completions with far fewer non-grid-aligned Gaussians.
Figure 5: Patchification with hierarchical Gaussians. A binary tree over the ground-truth Gaussians provides level-of-detail slices at each depth, allowing us to keep the Gaussian count low during training while preserving a faithful object representation. At the chosen depth, pairs of sibling nodes are merged into single transformer tokens, halving the sequence length while preserving spatial locality.
Objaverse
GSO
Method
#GS
PSNR ▲
DINO ▲
DS ▼
FID ▼
PSNR ▲
DINO ▲
DS ▼
FID ▼
⋆ ReconViaGen
500K
21.33
68.71
18.21
66.07
17.67
68.04
24.35
23.32
LGM
65K
22.88
66.46
17.05
88.43
19.65
58.01
26.41
54.96
LaRa
45K
27.79
71.72
12.39
93.19
25.64
60.20
22.28
69.18
GS-LRM
70K
27.90
77.57
10.33
65.06
27.57
72.20
16.11
34.13
Ours
8K
29.92
81.80
9.24
43.58
28.08
72.24
15.50
27.69
Table 1: Quantitative comparison of partially observed objects. We evaluate the case when four input views are concentrated in a hemisphere, leaving parts of the object unobserved. The values for DreamSim (DS) [ 12 ] and DINO similarity are ×100 . Our method shows strong performance on all metrics, with an order-of-magnitude fewer Gaussians (#GS). We highlight the metrics in blue, proportional to their percentile.
Figure 6: Full observation , Partial observation , and Validation cameras.
Objaverse
GSO
Method
#GS
PSNR ▲
SSIM ▲
LPIPS ▼
PSNR ▲
SSIM ▲
LPIPS ▼
⋆ ReconViaGen [ 2 ]
500K
22.54
94.01
82.27
18.29
90.42
102.3
LGM [ 49 ]
65K
26.65
95.50
57.71
23.64
92.35
73.61
LaRa [ 4 ]
45K
31.91
97.12
43.60
29.15
95.60
60.70
GS-LRM [ 69 ]
70K
31.03
97.10
32.80
31.13
95.00
30.93
Ours
8K
31.66
97.14
53.15
31.49
95.13
77.25
Table 2: Quantitative comparison of fully observed objects. We evaluate the case in which four input views provide full coverage of the object. The values for SSIM and LPIPS are ×100 and ×1000 , respectively. Our method achieves competitive quality with grid-aligned methods while generating an order-of-magnitude fewer Gaussians (#GS). We highlight the metrics in blue, proportional to their percentile.
Figure 7: Qualitative comparison on the Objaverse dataset [ 10 ] . The pixel-aligned LGM [ 49 ] and GS-LRM [ 69 ] produce severe artifacts and holes in unobserved regions, while LaRa [ 4 ] yields blurry completions. ReconViaGen’s generative reconstruction [ 2 ] synthesizes realistic content but is misaligned with the input views. Our method produces sharp, input-consistent reconstructions for both observed and unobserved regions.
Figure 8: Qualitative comparison on the GSO dataset. Additional examples comparing the input-fidelity and completeness of our method with the baselines.
Configuration
Full (PSNR ▲ )
Partial (PSNR ▲ )
Full method (Ours)
31.49
28.08
Training-time ablations
No Flow Matching
28.40 ( –3.09 )
26.30 ( –1.78 )
No MV Rendering Loss
29.66 ( –1.83 )
26.87 ( –1.21 )
No LoD Curriculum Training
30.00 ( –1.49 )
26.42 ( –1.66 )
Without Patching (4K GS)
30.47 ( –1.02 )
27.22 ( –0.86 )
Table 3: Ablation study. PSNR (dB) on GSO (1,030 objects) under full and partial observation, split into training-time and test-time ablations.
Figure 9: Qualitative Ablation results. Using a generative model as the core predictor, a timestep-weighted rendering loss, and coarse-to-fine curriculum training prove the most significant contributors to the quality of our final results.
# Views
PSNR ▲
2
28.22
4
31.49
6
32.04
9
32.39
Table 4: Number of input views. PSNR (dB) on GSO, full observations.