GS-STVSR achieves state-of-the-art continuous spatio-temporal video super-resolution quality with nearly constant inference time at standard scales and over 3x speedup at extreme scales using 2D Gaussian Splatting.
Sc-gs: Sparse-controlled gaussian splatting for editable dynamic scenes
7 Pith papers cite this work. Polarity classification is still indexing.
7
Pith papers citing it
citation-role summary
baseline 2
citation-polarity summary
fields
cs.CV 7roles
baseline 2polarities
baseline 2representative citing papers
PDR integrates a plug-and-play 2D Gaussian representation and CUDA renderer into pixel-based video predictors, replacing MSE loss with L1+SSIM to improve detail preservation on benchmarks like TaxiBJ and Human3.6M.
PersistGS decomposes scenes into per-object Gaussians and meshes, fits friction and velocity via differentiable simulation from pre-occlusion motion, and uses the resulting physics trajectory to position Gaussians during occlusions, improving PSNR by 2.46 dB over constant-velocity baselines on synth
Skelebones compresses 4D Gaussian shapes into compact, controllable bones and skeletons, delivering 17.3% PSNR gains over LBS and 21.7% over BoB for unseen poses while preserving reconstruction quality.
Structure-guided dynamic 3DGS methods deliver superior reconstruction fidelity and compactness on D-NeRF while gaussian-centric methods provide higher rendering speeds at the cost of quality variability and storage.
Dual-representation framework pairs fixed-topology meshes for physics with Gaussian splatting for rendering, but two conversion strategies from varying-topology reconstructions cause 65-80% geometric degradation and underperform native fixed-topology methods.
citing papers explorer
-
GS-STVSR: Ultra-Efficient Continuous Spatio-Temporal Video Super-Resolution via 2D Gaussian Splatting
GS-STVSR achieves state-of-the-art continuous spatio-temporal video super-resolution quality with nearly constant inference time at standard scales and over 3x speedup at extreme scales using 2D Gaussian Splatting.
-
Learning Video Dynamics with Predictive Differentiable Rendering
PDR integrates a plug-and-play 2D Gaussian representation and CUDA renderer into pixel-based video predictors, replacing MSE loss with L1+SSIM to improve detail preservation on benchmarks like TaxiBJ and Human3.6M.
-
PersistGS: Differentiable Physics for Object Permanence in 4D Gaussian Splatting
PersistGS decomposes scenes into per-object Gaussians and meshes, fits friction and velocity via differentiable simulation from pre-occlusion motion, and uses the resulting physics trajectory to position Gaussians during occlusions, improving PSNR by 2.46 dB over constant-velocity baselines on synth
-
GaussiAnimate: Reconstruct and Rig Animatable Categories with Level of Dynamics
Skelebones compresses 4D Gaussian shapes into compact, controllable bones and skeletons, delivering 17.3% PSNR gains over LBS and 21.7% over BoB for unseen poses while preserving reconstruction quality.
-
Beyond Static Gaussians: An Empirical Investigation of Architectural Paradigms for Dynamic 3D Scene Reconstruction
Structure-guided dynamic 3DGS methods deliver superior reconstruction fidelity and compactness on D-NeRF while gaussian-centric methods provide higher rendering speeds at the cost of quality variability and storage.
-
Real-Time Physics Simulation with Dynamic Mesh-Gaussian Reconstructions
Dual-representation framework pairs fixed-topology meshes for physics with Gaussian splatting for rendering, but two conversion strategies from varying-topology reconstructions cause 65-80% geometric degradation and underperform native fixed-topology methods.
- GSDeformer: Direct, Real-time and Extensible Cage-based Deformation for 3D Gaussian Splatting