InstantSplat: Sparse-view Gaussian Splatting in Seconds
Reviewed by Pith T0 review T1 audit T2 compute T3 formal T4 kernel pith:67PZIJUSrecord.jsonopen to challenge →
read the original abstract
While neural 3D reconstruction has advanced substantially, its performance significantly degrades with sparse-view data, which limits its broader applicability, since SfM is often unreliable in sparse-view scenarios where feature matches are scarce. In this paper, we introduce InstantSplat, a novel approach for addressing sparse-view 3D scene reconstruction at lightning-fast speed. InstantSplat employs a self-supervised framework that optimizes 3D scene representation and camera poses by unprojecting 2D pixels into 3D space and aligning them using differentiable neural rendering. The optimization process is initialized with a large-scale trained geometric foundation model, which provides dense priors that yield initial points through model inference, after which we further optimize all scene parameters using photometric errors. To mitigate redundancy introduced by the prior model, we propose a co-visibility-based geometry initialization, and a Gaussian-based bundle adjustment is employed to rapidly adapt both the scene representation and camera parameters without relying on a complex adaptive density control process. Overall, InstantSplat is compatible with multiple point-based representations for view synthesis and surface reconstruction. It achieves an acceleration of over 30x in reconstruction and improves visual quality (SSIM) from 0.3755 to 0.7624 compared to traditional SfM with 3D-GS.
This paper has not been read by Pith yet.
Forward citations
Cited by 26 Pith papers
-
Mind the Gap: Standard 3DGS Evaluation Primarily Measures Near-Trajectory Interpolation
Standard every-Nth-frame holdouts in 3D scene reconstruction primarily measure near-trajectory interpolation, with a consistent 3-12 dB gap to matched-count contiguous spatial holdouts that persists across Gaussian, n...
-
ZipSplat: Fewer Gaussians, Better Splats
ZipSplat uses multi-view token extraction followed by k-means clustering and attention to decode compact scene tokens into unconstrained 3D Gaussians, achieving SOTA pose-free results with ~6x fewer primitives.
-
PanoPlane: Plane-Aware Panoramic Completion for Sparse-View Indoor 3D Gaussian Splatting
PanoPlane achieves up to 17.8% PSNR gains in sparse-view indoor novel view synthesis by using training-free plane-aware panoramic completion to supervise 3D Gaussian Splatting.
-
Relightable Gaussian Splatting for Virtual Production Using Image-Based Illumination
A relightable Gaussian Splatting method for virtual production decomposes scenes into fixed appearance and variable lighting by parameterizing primitives to directly sample high-resolution background textures, enablin...
-
GSCompleter: A Distillation-Free Plugin for Metric-Aware 3D Gaussian Splatting Completion in Seconds
GSCompleter completes sparse 3D Gaussian Splatting scenes via a distillation-free generate-then-register pipeline using Stereo-Anchor lifting and Ray-Constrained Registration, delivering SOTA results on three benchmarks.
-
GSCompleter: A Distillation-Free Plugin for Metric-Aware 3D Gaussian Splatting Completion in Seconds
GSCompleter completes 3DGS scenes from sparse viewpoints using a generate-then-register workflow with stereo-anchor view selection and ray-constrained registration to achieve metric-aware results and SOTA performance ...
-
Splats in Splats++: Robust and Generalizable 3D Gaussian Splatting Steganography
Splats in Splats++ embeds messages into 3DGS via importance-graded SH encryption, hash-grid opacity mapping, and a gradient-gated consistency loss, achieving higher fidelity and robustness than prior methods.
-
Accelerating Transformer-Based Monocular SLAM via Geometric Utility Scoring
LeanGate is a lightweight feed-forward network that predicts geometric utility scores to skip over 90% of redundant frames in GFM-based monocular SLAM, reducing tracking FLOPs by 85% and achieving 5x speedup while mai...
-
DOC-GS: Dual-Domain Observation and Calibration for Reliable Sparse-View Gaussian Splatting
DOC-GS uses dual-domain calibration with continuous depth-guided dropout in optimization and dark channel prior evidence in observation to model and prune unreliable Gaussians, reducing haze and distortions in sparse-...
-
SparseCam4D: Spatio-Temporally Consistent 4D Reconstruction from Sparse Cameras
SparseCam4D achieves spatio-temporally consistent high-fidelity 4D reconstruction from sparse cameras via a Spatio-Temporal Distortion Field that corrects inconsistencies in generative observations.
-
Ocean4D: Generative Underwater 4D Reconstruction via Medium-Aware Video Diffusion
Ocean4D uses 4D geometrically consistent conditioning and a medium-aware denoising block in latent diffusion to generate cross-view consistent videos for underwater scenes from monocular input.
-
ForeSplat: Optimization-Aware Foresight for Feed-Forward 3D Gaussian Splatting
ForeSplat introduces MetaGrad, a meta-gradient training rule that makes feed-forward 3DGS predictions optimization-aware so they converge faster and to higher quality upon refinement.
-
ForeSplat: Optimization-Aware Foresight for Feed-Forward 3D Gaussian Splatting
ForeSplat introduces MetaGrad, a meta-gradient training rule that equips feed-forward 3DGS models to produce initializations optimized for rapid downstream refinement, improving convergence speed and final quality ove...
-
Generative 3D Gaussians with Learned Density Control
DeG models 3D Gaussians via learned octree density and uses VecSeq Sobol re-indexing to turn set generation into sequence modeling, claiming SOTA quality in single-image-to-3D.
-
TrackerSplat: Exploiting Point Tracking for Fast and Robust Dynamic 3D Gaussians Reconstruction
TrackerSplat pre-positions 3D Gaussians using point tracking trajectories to handle large inter-frame displacements in dynamic scene reconstruction.
-
UniRecGen: Unifying Multi-View 3D Reconstruction and Generation
UniRecGen unifies reconstruction and generation via shared canonical space and disentangled cooperative learning to produce complete, consistent 3D models from sparse views.
-
DAV-GSWT: Diffusion-Active-View Sampling for Data-Efficient Gaussian Splatting Wang Tiles
DAV-GSWT uses diffusion priors and active view sampling to synthesize high-fidelity Gaussian Splatting Wang Tiles from minimal observations while preserving visual quality and tile transitions.
-
DePT3R: Joint Dense Point Tracking and 3D Reconstruction of Dynamic Scenes in a Single Forward Pass
DePT3R performs joint dense point tracking and 3D reconstruction of dynamic scenes from multiple unposed images using a single neural network forward pass.
-
The Less You Depend, The More You Learn: Synthesizing Novel Views from Sparse, Unposed Images with Minimal 3D Knowledge
Data-centric novel view synthesis models with minimal 3D knowledge and no pose annotations scale better with data volume and outperform traditional bias-driven methods.
-
RoDyGS: Robust Dynamic Gaussian Splatting for Casual Videos
RoDyGS separates static and dynamic elements in monocular videos using Gaussian splatting with regularization and introduces the Kubric-MRig benchmark for pose-free dynamic novel view synthesis.
-
ViewCrafter: Taming Video Diffusion Models for High-fidelity Novel View Synthesis
ViewCrafter tames video diffusion models with point-based 3D guidance and iterative trajectory planning to produce high-fidelity novel views from single or sparse images.
-
VEPHand: View-Efficient Photometric Hand Performance Capture at Scale
End-to-end neural pipeline extracts hand geometry from unmasked limited-view images and registers it to a personalized tetrahedral model via volumetric offsets, achieving SOTA on over 12,000 sequences.
-
RoSplat: Robust Feed-Forward Pixel-wise Gaussian Splatting for Varying Input Views and High-Resolution Rendering
RoSplat adds alpha normalization for brightness consistency across varying input views and a 3D sampling regularizer to mitigate hole artifacts in high-resolution feed-forward Gaussian splatting.
-
3DTV: A Feedforward Interpolation Network for Real-Time View Synthesis
3DTV proposes a feedforward network for real-time sparse-view interpolation using Delaunay triplet selection and a pose-aware coarse-to-fine depth module, outperforming real-time baselines without scene-specific optimization.
-
The Role and Relationship of Initialization and Densification in 3D Gaussian Splatting
Current densification methods in 3D Gaussian Splatting do not significantly benefit from dense initializations and perform similarly to sparse SfM-based ones.
-
Turbo-GS: Accelerating 3D Gaussian Fitting for High-Quality Radiance Fields
Turbo-GS accelerates 3D Gaussian Splatting training via dilated rendering of pixel subsets, convergence-aware Gaussian budget allocation, and combined positional-appearance error densification to enable faster 4K fitt...
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.