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

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

Zooskool 8 Dogs In 1 Day Info

"Okay, let's do this," Jenny said to herself, taking a deep breath.

As the sun began to set on another busy day at Zooskool, Jenny reflected on the joy that these dogs brought to her life. She was already looking forward to the next day's adventures with her canine crew.

As the morning wore on, Jenny encountered more dogs: Daisy, the sweet-tempered Golden Retriever; Max, the mischievous Pug; Ginger, the spunky Chihuahua; Bear, the gentle Giant Schnauzer; and finally, there was Lola, the dramatic Poodle. Zooskool 8 Dogs In 1 Day

The third dog on the list was Luna, the elegant Greyhound. Luna was a gentle soul who loved to run, but only at her own pace. Jenny let Luna lead the way, enjoying the peaceful morning air as they glided through the neighborhood.

Next was Rocky, the rambunctious Boxer. Rocky was a ball of energy and loved to play tug-of-war with his favorite rope toy. Jenny was happy to oblige, laughing as Rocky bounded alongside her, his tail wagging wildly. "Okay, let's do this," Jenny said to herself,

First up was Bella, the playful Beagle. Bella loved to explore and sniff every fire hydrant and tree in sight. Jenny clipped on Bella's leash and they set off into the morning dew.

Each dog had their own unique personality and quirks, and Jenny loved getting to know them all. Despite the chaos of managing eight dogs in one day, Jenny wouldn't trade her job for anything. There was something special about spending time with these furry friends, and she felt grateful to be able to provide them with the exercise and attention they needed. As the morning wore on, Jenny encountered more

It was a typical Monday morning at Zooskool, the premier dog-walking and pet-sitting service in the city. Founder and lead dog wrangler, Jenny, was sipping her coffee and staring down at her schedule for the day. Her eyes widened as she took in the daunting list of eight dogs that needed to be walked that day.

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