FAST-GS: Frequency Aware Space-time Gaussian Splatting for Photorealistic Dynamic Novel View Synthesis
For researchers in dynamic scene reconstruction, this addresses the bottleneck of modeling complex high-frequency motions and long-term coherence in 4DGS, offering a practical improvement over existing single-polynomial approaches.
The paper proposes FAST-GS, a 4D Gaussian Splatting method that uses Fourier-based motion decomposition and a motion-aware regularization to improve dynamic novel view synthesis, achieving better accuracy and long-term stability on N3V and Google Immersive datasets.
4D Gaussian Splatting (4DGS) excels in dynamic 3D reconstruction and real-time novel view synthesis via efficient 4D Gaussian representations and parallelizable rendering. However, existing 4DGS approaches rely on a single polynomial to model motion, which limits performance in complex dynamic scenes where high-frequency motion components are prevalent, and fails to ensure long-term stability due to cumulative trajectory drift. To address these issues, we propose a Fourier Motion Modeling module: this paradigm decomposes motion into frequency-based sinusoidal components, capturing both low-frequency global trajectories and high-frequency local details to model complex motion patterns accurately. It retains the real-time rendering capability of 4DGS while improving complex motion fitting and long-term coherence. Additionally, we integrate a motion-aware regularization strategy into the loss function: it uses frequency-dependent weights to suppress high-frequency jitter while preserving low-frequency motion coherence. Extensive experiments on N3V and Google Immersive datasets from multiple scenarios demonstrate the effectiveness of our method.