MCGS-SLAM

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

Anonymous Author

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

Analysis of Single-Camera and Multi-Camera System

This experiment on the Waymo Open Dataset (Real World) demonstrates the effectiveness of our Multi-Camera Gaussian Splatting SLAM system. We evaluate the 3D mapping performance using three individual cameras, Front, Front-Left, and Front-Right, and compare these single-camera reconstructions against the Multi-Camera SLAM results.

The comparison highlights that the Multi-Camera SLAM leverages complementary viewpoints, providing more complete and geometrically consistent 3D reconstructions. In contrast, single-camera setups are prone to occlusions and limited fields of view, resulting in incomplete or distorted geometry. Our approach effectively fuses information from all three perspectives, achieving superior scene coverage and depth accuracy.

Right Image

Pola 2 < 2026 >

That night, Raya performed the penarikan —the withdrawal. She placed the mirror at the center of Pola Dua and whispered Kaleb’s forgotten name, learned from a century-old death record. As she spoke, the sand began to shimmer. A second shadow peeled off from her uncle’s sleeping form—grey, frayed at the edges, and humming with the sound of deep water.

She ran to Mbah Siti’s hut. The old woman was already waiting, holding a small mirror and a bowl of salt water.

It hesitated. Then it turned and walked into the mirror, spiraling inward until it vanished.

Raya shivered. “What happened?”

“He didn’t walk the second pattern,” Mbah Siti said. “Someone walked it for him. An echo of Kaleb. The sea doesn’t forget a broken promise.”

Old Mbah Siti was the last keeper of the second pattern. One evening, a curious teenager named Raya found her tracing invisible lines in the sand with a driftwood stick.

In the coastal village of Tanjung Harapan, the Pola was sacred. Every new moon, the fishermen would walk the spiral path carved into the eastern cliff—a living compass called Pola Satu (Pattern One). It was said that if you walked it barefoot before dawn, the sea would remember your name and grant you safe passage. pola 2

But no one spoke of Pola Dua .

The village doctor called it “parasomnia.” Mbah Siti called it bayangan terbelah —the divided shadow.

The next morning, Raya noticed something odd. Her uncle—a practical, unsuperstitious man—had started sleepwalking. Every night, he would rise from bed, walk to the eastern cliff, and trace an outward spiral before dawn. His eyes were open but empty. That night, Raya performed the penarikan —the withdrawal

She drew a shape that mirrored the cliff’s spiral—but inverted. Where Pola Satu curled inward like a nautilus, Pola Dua twisted outward like a storm unspooling.

Raya secretly filmed her uncle one night. When she reviewed the footage, her blood turned cold. In the recording, her uncle’s body walked Pola Satu —the safe spiral. But his shadow, stretched by moonlight, traced Pola Dua in reverse, pulling against his steps like a leash.

Her uncle woke gasping, his shadow normal once more. But Raya noticed something else: the mirror now held a faint, permanent spiral on its surface. And if she looked very closely, she could see a fisherman standing at its center, finally still, his two shadows rejoined. A second shadow peeled off from her uncle’s

“Long ago,” the old woman continued, “a fisherman named Kaleb grew tired of the sea’s silence. He wanted guarantees. So he walked Pola Dua at midnight—not to ask for safety, but to demand a catch.”


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