LoCoSplat: Real-Time Feed-Forward 3D Gaussian Splatting with Minimal 3D Reasoning

Sinan Wang; Jinjin He; Yuchen Sun; Duowen Chen; Shenyifan Lu; Bo Zhu

cs.CVOct 3, 2026

LoCoSplat: Real-Time Feed-Forward 3D Gaussian Splatting with Minimal 3D Reasoning

Authors: Sinan Wang, Jinjin He, Yuchen Sun, Duowen Chen, Shenyifan Lu, Bo Zhu

Organizations: Georgia Institute of Technology

Abstract

Feed-forward 3D Gaussian Splatting (3DGS) increasingly aggregates multi-view evidence with heavy learned 3D networks. We propose LoCoSplat (Local-Context Splatting), motivated by the observation that a Gaussian is a local primitive: once depth is predicted, what the 3D stage must add (scale, rotation, opacity) depends on the point cloud around each anchor, and a fixed local average of that neighbourhood is enough to supply it, no heavy network required. LoCoSplat realises exactly this average: it splats a 16-d linear projection of the point features into a fine and a coarse grid and reads both back at each anchor with a 0.14M-parameter pointwise MLP; with no learned 3D network and no dynamic sparse computation, its whole encoder runs as one fp16 CUDA graph. On RealEstate10K, LoCoSplat outperforms every prior feed-forward method on PSNR, SSIM, and LPIPS at 6, 12, and 24 views, with a margin that widens as views densify (+3.3 PSNR over VolSplat, the prior voxel-aligned state of the art, at 24 views) and grows further under zero-shot transfer to ACID and fine-tuning on ScanNet. It reconstructs a 6-view scene in 33 ms on one NVIDIA RTX PRO 6000 GPU, the fastest of seven feed-forward methods and $4.2\times$ faster than the previous state of the art, trains $2.7\times$ faster ( $5.9\times$ at 24 views), and uses $6.7\times$ less inference memory.

Figures & tables

Figure 1: LoCoSplat overview. Predicted depth and per-pixel features are unprojected and voxelized into anchors; the 3D stage averages anchor features in a fine and a coarse grid and decodes each anchor pointwise. The whole encoder runs as one CUDA graph.

Figure 2: Local context as a fixed average. Points in the coarse cell $C_{H}(a)$ are projected by ${\bm{W}}_{1}$ and averaged, with no parameters, into ${\bm{u}}_{1}(a)$ ; the anchor’s feature gives ${\bm{u}}_{0}(a)={\bm{W}}_{0}{\bm{f}}_{a}$ . The pointwise map $g$ fuses $[{\bm{u}}_{0};{\bm{u}}_{1};{\bm{f}}_{a}]$ residually and the head $h$ decodes $Y_{a}$ ( Eqs. 3 , 4 and 5 ).

Method	6 views			12 views			24 views
Method	PSNR $\uparrow$	SSIM $\uparrow$	LPIPS $\downarrow$	PSNR $\uparrow$	SSIM $\uparrow$	LPIPS $\downarrow$	PSNR $\uparrow$	SSIM $\uparrow$	LPIPS $\downarrow$
pixelSplat ( Charatan et al., 2024 )	28.95	0.900	0.163	OOM	OOM	OOM	OOM	OOM	OOM
MVSplat ( Chen et al., 2024 )	29.13	0.924	0.091	26.97	0.912	0.101	26.23	0.903	0.108
TranSplat ( Zhang et al., 2025 )	29.62	0.928	0.084	28.00	0.920	0.089	26.65	0.884	0.115
DepthSplat ( Xu et al., 2025 )	30.52	0.931	0.079	28.54	0.919	0.088	26.26	0.880	0.115
GGN ( Zhang et al., 2024 )	26.68	0.879	0.133	25.83	0.870	0.142	23.86	0.840	0.171

Table 1: RealEstate10K ( Zhou et al., 2018 ) , 6/12/24 views. pixelSplat–GGN are quoted from VolSplat’s retrainings ( Wang et al., 2026a ) ; our retraining uses VolSplat’s official protocol ( Section 3.1 ). ‡ Pose-free, released checkpoints with ground-truth cameras (different training corpora and resolutions, no retraining). OOM: cannot infer on 96 GB.

Figure 3: RealEstate10K ( Zhou et al., 2018 ) , 6 input views. LoCoSplat (“Ours”) preserves fine detail and thin structures with less ghosting; red boxes highlight local differences. Checkpoints follow Table 1 (VolSplat: our retraining; AnySplat and WorldMirror: released checkpoints with ground-truth cameras).

Method	PSNR $\uparrow$	Params train. / total	Inference (ms)	Recon. FPS	Peak mem. (MB)	Train (it/s)	Train wall-clock
pixelSplat	28.95	119.0 / 119.0 M	238.5	4.2	31,869	1.05	$\sim$ 39.6 h
MVSplat	29.13	12.0 / 12.0 M	93.8	10.7	3,528	2.92	$\sim$ 14.3 h
DepthSplat	30.52	120.7 / 120.7 M	81.2	12.3	7,204	2.44	$\sim$ 17.1 h
GGN	26.68	12.1 / 12.1 M	177.0	5.6	3,489	2.25	$\sim$ 18.5 h
AnySplat ‡	25.37	886.4 / 1,190.7 M	85.5	11.7	3,774	—	—
WorldMirror ‡	29.20	1,263.3 / 1,263.3 M	90.8	11.0	5,673	—	—

Table 2: Six-view reconstruction efficiency (RealEstate10K; same GPU and timing protocol for all methods, Appendix A ). Baselines run in plain fp32 as released; LoCoSplat on its deployed path (fp16, CUDA graph) and in plain fp32. Methods as in Table 1 . Training: 4 GPUs, batch 1/GPU, wall-clock extrapolated from the measured step rate to 150k steps. ‡ Pose-free, released defaults, not retrainable here.

Figure 4: Zero-shot ACID ( Liu et al., 2021 ) , 6 input views. Models trained on RealEstate10K, checkpoints as in Table 4 ; LoCoSplat preserves thin structures and road markings with less blur and ghosting (red boxes).

Table 7

Figure 5: ScanNet ( Dai et al., 2017 ) , 6 input views. LoCoSplat and VolSplat are fine-tuned under VolSplat’s protocol, the other methods use released checkpoints ( Table 4 ); LoCoSplat preserves shelves and rug patterns with fewer artifacts.

Model	$D$	$\Delta u$ (rel.)	PSNR $\uparrow$	SSIM $\uparrow$	LPIPS $\downarrow$	Paired $\Delta$ PSNR vs. LoCoSplat	LoCoSplat better
12 views
LoCoSplat, $D{=}256$ (pre-reg.)	256	0.23	28.39	0.895	0.109	$-3.10$	99%
LoCoSplat	64	0.92	31.49	0.943	0.073	—	—
24 views
VolSplat (retrained)	128	0.22	27.24	0.881	0.127	$-3.3$	98%
LoCoSplat, $D{=}128$	128	0.22	28.08	0.888	0.114	$-2.49$	98%

Table 5: Plane-sweep candidate count and the dense-view collapse. Full-budget retrainings of LoCoSplat with one component changed.

\Delta u

: adjacent-candidate disparity relative to 6 views,

D{=}128

( Eq. 7 ); paired

\Delta

PSNR vs. LoCoSplat at the same view count.

D{=}256

and

D{=}32

were pre-registered predictions; the 12-view reference is re-evaluated for this table (31.49 vs. 31.47 in Table 1 ).

Figure 6: 24-view training loss (5k-step mean). Both

D{=}128

models plateau at twice the loss; the original VolSplat with

D{=}64

follows LoCoSplat. Dots: PSNR on the 24-view index at 25k and 50k steps. Curves cover the first 90k steps of the 150k-step schedule.

Appendix figures & tables13 assets

Supplementary material from the paper’s appendix.

Appendix

Method	Encoder (ms) $\downarrow$			Encoder-only FPS $\uparrow$			Encoder + render, FPS $\uparrow$
Method	6v	12v	24v	6v	12v	24v	6v	12v	24v
pixelSplat	238.5	—	—	4.2	—	—	4.0	—	—
MVSplat	93.8	303.9	1111.7	10.7	3.3	0.9	9.6	3.1	0.9
DepthSplat	81.2	149.8	306.4	12.3	6.7	3.3	11.1	6.1	3.0
GGN	177.0	483.8	1451.7	5.6	2.1	0.7	5.4	2.0	0.7
AnySplat ‡	85.5	116.6	209.0	11.7	8.6	4.8	10.5	7.6	4.3

Appendix

Table 6: Reconstruction speed, 6/12/24 views (RealEstate10K, 100-scene medians; protocol above). FPS counts scenes; “+ render” includes 8 targets. Checkpoints: MVSplat, DepthSplat, and GGN are timed at 12 and 24 views with their released 6-view checkpoints (their cost is shape-only, so timing is unaffected); VolSplat at 24 views with our 12-view checkpoint ( § ); LoCoSplat with its own 6-, 12-, and 24-view checkpoints. Baselines run in plain fp32 as released, LoCoSplat on its deployed fp16 CUDA-graph path (fp32 eager: 56.9 ms at 6 views, the same ordering). ‡ Pose-free, released defaults:

252^{2}

, bf16 backbone, multi-dataset pretraining; WorldMirror uses ground-truth camera priors.

Method	Render 8 views (ms) $\downarrow$			Peak memory (MB) $\downarrow$			#Gaussians (mean)
Method	6v	12v	24v	6v	12v	24v	6v	12v	24v
pixelSplat	10.4	—	—	31,869	—	—	393,216	—	—
MVSplat	9.9	17.2	31.7	3,528	7,277	24,341	393,216	786,432	1,572,864
DepthSplat	8.4	13.4	23.8	7,204	13,912	27,330	393,216	786,432	1,572,864
GGN	9.2	9.3	9.8	3,489	7,158	24,010	176,837	238,445	319,885
AnySplat ‡	10.4	14.9	23.8	3,774	4,199	5,106	343,347	640,670	1,145,219

Appendix

Table 7: Rendering, memory, and Gaussian counts. Protocol and ‡ as in Table 6 .

Configuration	Encoder (ms)	PSNR
Two-grid model ( Section 3.5 ), original code	94.4	31.63
+ fused ops (SDPA, packed voxel unique, fused gathers)	87.4	31.63
+ fp16 image-side autocast	83.8	31.63
+ runtime type-checking removed	74.4	31.63
+ CUDA-graphed 2D image side	67.4	31.63
graph-native retrain (pointwise head, fine-from-base, $D{=}64$ )	52.0	31.60

Appendix

Table 8: Cumulative speedups. Top: checkpoint-compatible optimizations; bottom: retrained models with deployed stacks.

Figure 7: RealEstate10K ( Zhou et al., 2018 ) , 12 input views (four shown). LoCoSplat (“Ours”) preserves masonry, branches, and distant structures where baselines blur or introduce floaters. Red boxes highlight local differences. VolSplat is retrained at 12 views; pose-free baselines use released checkpoints with ground-truth cameras, as in Fig. 3 .

Figure 8: RealEstate10K ( Zhou et al., 2018 ) , 24 input views (four shown). In these selected examples, LoCoSplat (“Ours”) preserves branches, text, and fine structure where VolSplat streaks and pose-free baselines blur. Red boxes highlight local differences. Checkpoints, evaluation paths, and example selection are detailed below.

Stage	PSNR $\uparrow$	LPIPS $\downarrow$	Enc. (ms)
VolSplat (sparse 3D U-Net, 17M)	30.93	0.078	137.0
$\to$ 8-level grid + MLP	31.64	0.073	118
$\to$ 2 levels $\{0.001,0.128\}$	31.63	0.072	94.4
$\to$ pointwise head, $D{=}64$	31.60	0.073	58.9
$\to$ ViT-B/narrow iface, DWConv (= LoCoSplat)	31.51	0.074	33.0

Appendix

Table 9: Cumulative evolution to LoCoSplat (150k steps, RealEstate10K, 6 views, 709 scenes). Enc.: encoder time on one RTX PRO 6000; intermediate models run their original code, LoCoSplat its full inference stack ( Table 8 ).

Figure 9: More qualitative comparisons on RealEstate10K ( Zhou et al., 2018 ) , 6 input views (four shown). Checkpoints and evaluation follow Fig. 3 ; selection is described in Section E.4 . Red boxes highlight local differences.

Model (12 views)	3D stage	PSNR $\uparrow$	SSIM $\uparrow$	LPIPS $\downarrow$	$\Delta$ PSNR
VolSplat (retrained)	sparse 3D U-Net	30.86	0.939	0.077	—
LoCoSplat	two-level grid + MLP	31.47	0.942	0.073	$\mathbf{+0.61}$ vs. retrained (paired)

Appendix

Table 10: LoCoSplat vs. VolSplat at 12 views. RealEstate10K, 12 input views. Our retraining of VolSplat and LoCoSplat (which also carries the backbone-side changes of Section 2.1 ) on the official 12-view index;

\Delta

is the per-scene paired difference against the retrained VolSplat. The full LoCoSplat pipeline, whose 3D stage is the two-level grid and a pointwise MLP, gains

+0.61

, a whole-pipeline gain (the isolated 3D-stage comparison is Table 9 at 6 views). Our 12-view retraining (30.86) is stronger than the published model (29.40, Section B.2 ).

Coarse source	Learned?	PSNR $\uparrow$	Paired $\Delta$ vs. ref.
Learned re-splat (reference, 2 runs, same seed)	✓	31.60 / 31.63	—
Fine field, hard box pooling	$\times$	31.58	$-0.013$ (tie)
Fine field, Gaussian low-pass	$\times$	31.59	$-0.005$ (tie)
Fine field, anchor-centered radius mean	$\times$	31.53	$-0.062$
Geometry-only statistics (no appearance)	$\times$	31.63	$+0.030$ (tie)

Appendix

Table 11: Coarse-source ablation on the two-grid model ( Section 3.5 : VolSplat’s 164-d features, trilinear grids, sparse-convolutional head,

D{=}128

; full 150k budget; RealEstate10K, 6 views, 709 scenes). Its reference is therefore the two-grid model (31.60 / 31.63), not LoCoSplat (31.51, Table 12 ). The reference row (learned re-splat of the 16-d projected features) is the coarse level LoCoSplat ships. Paired

\Delta

PSNR vs. the reference run. The reference row lists two training runs at the same seed (this code tree / earlier run); three training seeds of the two-grid model lie pairwise within

0.03

PSNR ( Section B.2 ), which bounds the replicate noise (three draws of LoCoSplat give

0.02

, Table 12 ), and differences within it are marked (tie).

Config	Scales	PSNR $\uparrow$	SSIM $\uparrow$	LPIPS $\downarrow$	$\Delta$ PSNR vs. LoCoSplat
1 level (fine only)	$0.001$	31.38	0.941	0.075	$-0.13$
2 levels (LoCoSplat)	$\{0.001,0.128\}$	31.51	0.942	0.074	—
2 levels (LoCoSplat, second seed)	$\{0.001,0.128\}$	31.50	0.941	0.074	$-0.01$ (tie)
2 levels (LoCoSplat, third seed)	$\{0.001,0.128\}$	31.50	0.941	0.074	$-0.01$ (tie)
4 levels ( $\times 2$ )	$0.001$ – $0.008$	31.50	0.941	0.074	$-0.01$ (tie)
8 levels ( $\times 2$ )	$0.001$ – $0.128$	31.49	0.941	0.074	$-0.02$ (tie)

Appendix

Table 12: Level count vs. scale range on the final LoCoSplat configuration at full budget (only the grid levels change; RealEstate10K, 6 views, 709 scenes; the reference is LoCoSplat itself, 31.51, unlike Table 11 ).

\Delta

PSNR is the difference of the 709-scene means vs. LoCoSplat, which on a common scene set equals the paired mean. (tie): within replicate noise, bounded by three seed draws of LoCoSplat (rows 2–4: largest paired difference

0.02

PSNR, sample s.d.

<0.01

) and by three seed draws of the two-grid model of Table 11 (

\sim

0.03). Removing the coarse levels is the one change that hurts (1 level vs. 4 / 8 levels, paired:

-0.123

-0.113

); four narrow levels, eight wide levels in two seed draws, and the two levels of LoCoSplat in three seed draws tie (8 vs. 4, paired:

-0.010

; 8 vs. 2 at the same seed:

-0.006

); extending the coarsest scale from

0.128

2.048

is the only change that helps.

Figure 10: More qualitative comparisons on RealEstate10K ( Zhou et al., 2018 ) , 24 input views (four shown). Checkpoints and evaluation follow Fig. 8 ; selection is described in Section E.4 . Red boxes highlight local differences.

Model	Views	$D$	$\Delta u$ (px)	$H/\log D$	$w_{90}$ (bins)	Modal-bin share	MV depth err.	Status
LoCoSplat	6	64	0.110	0.42	7	13%	1.8%	healthy
LoCoSplat	12	64	0.050	0.51	8	14%	2.0%	healthy
LoCoSplat, $D{=}256$	12	256	0.012	0.000	1	100%	48%	collapsed
LoCoSplat, $D{=}32$	24	32	0.047	0.51	5	19%	1.5%	healthy
LoCoSplat	24	64	0.023	0.47	8	14%	3.2%	healthy
LoCoSplat, $D{=}128$	24	128	0.012	0.000	1	100%	46%	collapsed

Appendix

Table 13: The depth branch of healthy and collapsed models.

\Delta u

: exact median per-candidate disparity in scale-0 feature pixels over the full official index (709/696/680 scenes). Remaining columns: scale-0 matching distribution on the first 50 scenes of the index, pooled over all pixels: normalized entropy

H/\log D

(median),

w_{90}

= bins around the argmax holding 90% of the mass (median), share of pixels whose argmax is the most frequent bin, and median relative error of the multi-view depth against the final fused depth. Collapsed models are input-independent one-hots at a fixed candidate; at initialization every configuration is broad and alike. Two-grid model: Section 3.5 .

Figure 11: More qualitative comparisons on RealEstate10K ( Zhou et al., 2018 ) , 12 input views (four shown). Checkpoints and evaluation follow Fig. 7 ; selection is described in Section E.4 . Red boxes highlight local differences.

Explore similar work

Apr 16, 2026cs.CV

GlobalSplat: Efficient Feed-Forward 3D Gaussian Splatting via Global Scene Tokens

The efficient spatial allocation of primitives serves as the foundation of 3D Gaussian Splatting, as it directly dictates the synergy between representation compactness, reconstruction speed, and rendering fidelity. Previous solutions, whether based on iterative optimization or feed-forward inference, suffer from significant trade-offs between these goals, mainly due to the reliance on local, heuristic-driven allocation strategies that lack global scene awareness. Specifically, current feed-forward methods are largely pixel-aligned or voxel-aligned. By unprojecting pixels into dense, view-aligned primitives, they bake redundancy into the 3D asset. As more input views are added, the representation size increases and global consistency becomes fragile. To this end, we introduce GlobalSplat, a framework built on the principle of align first, decode later. Our approach learns a compact, global, latent scene representation that encodes multi-view input and resolves cross-view correspondences before decoding any explicit 3D geometry. Crucially, this formulation enables compact, globally consistent reconstructions without relying on pretrained pixel-prediction backbones or reusing latent features from dense baselines. Utilizing a coarse-to-fine training curriculum that gradually increases decoded capacity, GlobalSplat natively prevents representation bloat. On RealEstate10K and ACID, our model achieves competitive novel-view synthesis performance while utilizing as few as 16K Gaussians, significantly less than required by dense pipelines, obtaining a light 4MB footprint. Further, GlobalSplat enables significantly faster inference than the baselines, operating under 78 milliseconds in a single forward pass. Project page is available at https://r-itk.github.io/globalsplat/

Roni Itkin, Noam Issachar, Yehonatan Keypur +3

The Hebrew University of Jerusalem · Westlake University

3D Gaussian Splatting 3DGS Acceleration Novel View Synthesis 3D Scene Representation Feed-Forward 3D Reconstruction

Ask PDF

Sep 27, 2026cs.CV

FeCoSplat: Feedback-Guided Compression for Feed-Forward 3D Gaussian Splatting

Feed-forward 3D Gaussian Splatting (3DGS) enables efficient novel-view synthesis from sparse multi-view images, yet its representations remain costly to store and transmit. Existing approaches compress either the input images, incurring heavy receiver-side reconstruction, or the reconstructed Gaussian primitives, which are difficult to compress due to their heterogeneous and irregular attributes. We instead compress compact intermediate features, providing a better balance between compression efficiency and receiver-side complexity. Based on this paradigm, we propose FeCoSplat, a feedback-guided compression framework for feed-forward 3DGS. FeCoSplat first compresses multi-view features to obtain an intermediate 3DGS, whose rendered views are used as feedback to guide a second-stage compression for further refinement. The resulting bitstreams are decoded into a compact implicit state, from which the final Gaussian primitives are reconstructed with a lightweight predictor. Experiments demonstrate that FeCoSplat achieves favorable rate--distortion performance, particularly at low bitrates, while requiring only 3.45M parameters for receiver-side Gaussian reconstruction. Code will be released soon.

Yuxuan Li, Yihang Chen, Yufeng Zhang +2

Shanghai Jiao Tong University · Monash University

3DGS Compression 3D Gaussian Splatting Feed-Forward 3D Reconstruction

Ask PDF

Jun 3, 2026cs.CV

ZipSplat: Fewer Gaussians, Better Splats

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 https://veichta.com/zipsplat.

Alexander Veicht, Sunghwan Hong, Dániel Baráth +1

ETH Zürich · Microsoft

3D Gaussian Splatting Multi-View 3D Reconstruction Feed-Forward 3D Reconstruction

Ask PDF