GaussianVideo: Efficient Video Representation via Hierarchical Gaussian Splatting
This work addresses memory usage, training time, and temporal consistency issues in video representation for applications like compression and simulations, representing a novel method for a known bottleneck.
The paper tackled the problem of inefficient neural video representations by introducing a method combining 3D Gaussian splatting with continuous camera motion modeling, achieving state-of-the-art performance with high-quality rendering and strong temporal consistency across diverse video datasets.
Efficient neural representations for dynamic video scenes are critical for applications ranging from video compression to interactive simulations. Yet, existing methods often face challenges related to high memory usage, lengthy training times, and temporal consistency. To address these issues, we introduce a novel neural video representation that combines 3D Gaussian splatting with continuous camera motion modeling. By leveraging Neural ODEs, our approach learns smooth camera trajectories while maintaining an explicit 3D scene representation through Gaussians. Additionally, we introduce a spatiotemporal hierarchical learning strategy, progressively refining spatial and temporal features to enhance reconstruction quality and accelerate convergence. This memory-efficient approach achieves high-quality rendering at impressive speeds. Experimental results show that our hierarchical learning, combined with robust camera motion modeling, captures complex dynamic scenes with strong temporal consistency, achieving state-of-the-art performance across diverse video datasets in both high- and low-motion scenarios.