MCGS-SLAM

A Multi-Camera SLAM Framework Using Gaussian Splatting for High-Fidelity Mapping

Anonymous Author
Code and data [Available Upon Acceptance]

SLAM System Pipeline

Our method performs real-time SLAM by fusing synchronized inputs from a multi-camera rig into a unified 3D Gaussian map. It first selects keyframes and estimates depth and normal maps for each camera, then jointly optimizes poses and depths via multi-camera bundle adjustment and scale-consistent depth alignment. Refined keyframes are fused into a dense Gaussian map using differentiable rasterization, interleaved with densification and pruning. An optional offline stage further refines camera trajectories and map quality. The system supports RGB inputs, enabling accurate tracking and photorealistic reconstruction.

Right Image

(2022)(1)(1)(2).mp4 — Friends Uncut

At any moment in the video, you can toggle between the broadcast view, a wide-lens "stage" view showing the crew/audience, or a specific "isolated" camera on one actor to see their reactions when they aren't the focus of the main shot.

Since "Uncut" implies raw, behind-the-scenes, or extended footage, this feature would allow viewers to:

A subtle, non-intrusive icon that pops up when a scene was improvised or had multiple takes. Clicking it displays a quick text snippet or a split-screen showing the alternative versions that didn't make the final cut.

Analysis of Single-Camera and Multi-Camera SLAM (Mapping)

At any moment in the video, you can toggle between the broadcast view, a wide-lens "stage" view showing the crew/audience, or a specific "isolated" camera on one actor to see their reactions when they aren't the focus of the main shot.

Since "Uncut" implies raw, behind-the-scenes, or extended footage, this feature would allow viewers to:

A subtle, non-intrusive icon that pops up when a scene was improvised or had multiple takes. Clicking it displays a quick text snippet or a split-screen showing the alternative versions that didn't make the final cut.

Analysis of Single-Camera and Multi-Camera SLAM (Tracking)

In this section, we benchmark tracking accuracy across eight driving sequences from the Waymo dataset (Real World). MCGS-SLAM achieves the lowest average ATE, significantly outperforming single-camera methods.
Right Image

We further evaluate tracking on four sequences from the Oxford Spires dataset (Real World). MCGS-SLAM consistently yields the best performance, demonstrating robust trajectory estimation in large-scale outdoor environments.
Right Image

Right Image