Search papers, labs, and topics across Lattice.
This paper introduces CaT-GS, an innovative rendering pipeline that enhances 3D Gaussian Splatting (3DGS) for large-scale scenes by addressing inefficiencies in GPU utilization. By implementing a speculative multi-frame preprocessing method and an inter-frame caching mechanism, the authors significantly reduce redundant computations and viewpoint-based rendering stages. The results show that CaT-GS achieves up to a 10x speedup compared to the original 3DGS and up to 70% over previous state-of-the-art methods, setting a new benchmark for real-time rendering fidelity.
Achieving up to 10x faster rendering speeds, CaT-GS redefines efficiency in 3D Gaussian Splatting for large-scale scenes.
Recent breakthroughs in 3D Gaussian Splatting (3DGS) have advanced neural rendering with high fidelity and speed. However, its performance degrades significantly in large-scale scenes due to the computational burden of tile-based rasterization. Existing optimization efforts either require costly scene re-training or focus on narrow aspects of the pipeline, overlooking critical inefficiencies in real-world deployments. Through a comprehensive analysis, we identify three primary sources of redundancy and low GPU utilization: redundant inter-frame pre-processing, viewpoint-based occlusion redundancy, and severe tile-level load imbalance. To address these issues, we propose CaT-GS, a novel and efficient 3DGS rendering pipeline. CaT-GS introduces a speculative multi-frame preprocessing method to eliminate redundant computations across consecutive frames, and an inter-frame caching mechanism to eliminate viewpoint redundant rendering stages. Furthermore, it refactors rasterization tasks with a dedicated kernel to mitigate tile load imbalance, significantly boosting GPU utilization. Extensive experiments demonstrate that CaT-GS achieves a speedup of up to 10 times over the original 3DGS and up to 70% over previous state-of-the-art methods, establishing a new benchmark for high-fidelity, real-time rendering of large-scale scenes.
