GETA-3DGS is the first automatic joint structured pruning and quantization framework for 3D Gaussian Splatting, achieving roughly 5x storage reduction on standard datasets without per-scene thresholds.
Compact3d: Smaller and faster gaussian splatting with vector quanti- zation
7 Pith papers cite this work. Polarity classification is still indexing.
7
Pith papers citing it
citation-role summary
background 1
method 1
citation-polarity summary
representative citing papers
GEMM-GS converts 3DGS blending into GEMM form to use Tensor Cores, yielding 1.42x speedup over vanilla 3DGS and further gains when stacked with prior accelerators.
SIG synchronizes training-image resolution and Gaussian densification to measured scene bandwidth, plus sphere constraints on primitives, delivering better PSNR and 1.4–1.5× per-block speedups on large outdoor scenes.
Flow Splatting extends 4D Gaussian volumes with time-varying means and covariances, approximates a velocity field, and splats it to render optical flow for supervising dynamic reconstruction from monocular video.
Presents GScomp-QA dataset with subjective perceptual scores for compressed Gaussian Splatting, enabling rate-distortion analysis and showing that 18 objective metrics fail to fully capture GS-specific distortions.
POTR introduces simultaneous-effect pruning via a modified 3DGS rasterizer and entropy-reducing lighting coefficient recomputation to outperform prior post-training 3DGS compression methods in rate-distortion and inference speed.
A survey compiling principles, applications, benchmarks, and challenges of 3D Gaussian Splatting for explicit 3D scene representation.
citing papers explorer
-
GETA-3DGS: Automatic Joint Structured Pruning and Quantization for 3D Gaussian Splatting
GETA-3DGS is the first automatic joint structured pruning and quantization framework for 3D Gaussian Splatting, achieving roughly 5x storage reduction on standard datasets without per-scene thresholds.
-
GEMM-GS: Accelerating 3D Gaussian Splatting on Tensor Cores with GEMM-Compatible Blending
GEMM-GS converts 3DGS blending into GEMM form to use Tensor Cores, yielding 1.42x speedup over vanilla 3DGS and further gains when stacked with prior accelerators.
-
Signal Structure-Aware Gaussian Splatting for Large-Scale Scene Reconstruction
SIG synchronizes training-image resolution and Gaussian densification to measured scene bandwidth, plus sphere constraints on primitives, delivering better PSNR and 1.4–1.5× per-block speedups on large outdoor scenes.
-
Learning Efficient 4D Gaussian Representations from Monocular Videos with Flow Splatting
Flow Splatting extends 4D Gaussian volumes with time-varying means and covariances, approximates a velocity field, and splats it to render optical flow for supervising dynamic reconstruction from monocular video.
-
GScomp-QA: A Subjective Dataset for Quality Assessment of Compressed Gaussian Splatting
Presents GScomp-QA dataset with subjective perceptual scores for compressed Gaussian Splatting, enabling rate-distortion analysis and showing that 18 objective metrics fail to fully capture GS-specific distortions.
-
POTR: Post-Training 3DGS Compression
POTR introduces simultaneous-effect pruning via a modified 3DGS rasterizer and entropy-reducing lighting coefficient recomputation to outperform prior post-training 3DGS compression methods in rate-distortion and inference speed.
-
A Survey on 3D Gaussian Splatting
A survey compiling principles, applications, benchmarks, and challenges of 3D Gaussian Splatting for explicit 3D scene representation.