Papers
Sign in to view your remaining parses.
Tag Filter
3D Gaussian Splatting representation
GaussianObject: High-Quality 3D Object Reconstruction from Four Views
with Gaussian Splatting
Published:2/16/2024
3D Gaussian Splatting representationSparse-View 3D ReconstructionDiffusion-Based 3D RepairMulti-View Consistency OptimizationCOLMAP-Free Camera Pose Estimation
GaussianObject uses Gaussian splatting with visual hull and floater elimination to reconstruct highquality 3D objects from only four views. A diffusionbased Gaussian repair model restores missing details, and a COLMAPfree variant enables posefree reconstruction, outperforming
03
G4Splat: Geometry-Guided Gaussian Splatting with Generative Prior
Published:10/14/2025
3D Gaussian Splatting representationGenerative PriorMulti-View Consistency ModelingDiffusion Model-Based 3D ReconstructionGeometry-Guided Planar Structures
G4Splat integrates geometryguided depth supervision with generative priors to improve 3D scene reconstruction quality and multiview consistency, outperforming baselines on diverse datasets and supporting singleview and unposed video inputs.
06
Difix3D+: Improving 3D Reconstructions with Single-Step Diffusion Models
Published:3/4/2025
Single-Step Diffusion Model for 3D Reconstruction EnhancementNeural Radiance Fields (NeRF)3D Gaussian Splatting representationNovel-View SynthesisPseudo-Training View Refinement for 3D Reconstruction
Difix3D uses a singlestep diffusion model, Difix, to enhance 3D reconstructions by removing artifacts, refining pseudotraining views, and distilling improvements back into 3D, achieving superior novelview synthesis and doubled average FID scores across NeRF and 3DGS represent
03
ManiGaussian: Dynamic Gaussian Splatting for Multi-task Robotic
Manipulation
Published:3/13/2024
3D Gaussian Splatting representationMultimodal Robot LearningMulti-Task Robotic ManipulationFuture Scene ReconstructionDynamic Semantic Propagation
ManiGaussian uses dynamic Gaussian splatting and future scene reconstruction to capture spatiotemporal dynamics for multitask robotic manipulation, outperforming stateoftheart methods by 13.1% in success rate on RLBench benchmarks.
05