Search papers, labs, and topics across Lattice.
This paper introduces a multimodal 3D scene rendering framework that integrates radio-frequency (RF) sensing with 3D Gaussian Splatting (GS) to improve robustness and efficiency, especially in challenging visual conditions. They leverage RF data, specifically from automotive radar, to generate a high-quality 3D point cloud for initializing Gaussian primitives, addressing limitations of vision-only GS pipelines that require many camera views or struggle with adverse conditions. Experiments demonstrate that incorporating RF sensing into GS pipelines leads to high-fidelity 3D scene rendering with improved structural accuracy.
RF sensing can rescue 3D Gaussian Splatting from the visual data graveyard of adverse weather, low light, and occlusions.
3D scene reconstruction and rendering are core tasks in computer vision, with applications spanning industrial monitoring, robotics, and autonomous driving. Recent advances in 3D Gaussian Splatting (GS) and its variants have achieved impressive rendering fidelity while maintaining high computational and memory efficiency. However, conventional vision-based GS pipelines typically rely on a sufficient number of camera views to initialize the Gaussian primitives and train their parameters, typically incurring additional processing cost during initialization while falling short in conditions where visual cues are unreliable, such as adverse weather, low illumination, or partial occlusions. To cope with these challenges, and motivated by the robustness of radio-frequency (RF) signals to weather, lighting, and occlusions, we introduce a multimodal framework that integrates RF sensing, such as automotive radar, with GS-based rendering as a more efficient and robust alternative to vision-only GS rendering. The proposed approach enables efficient depth prediction from only sparse RF-based depth measurements, yielding a high-quality 3D point cloud for initializing Gaussian functions across diverse GS architectures. Numerical tests demonstrate the merits of judiciously incorporating RF sensing into GS pipelines, achieving high-fidelity 3D scene rendering driven by RF-informed structural accuracy.
