# DyPho-SLAM : Real-time Photorealistic SLAM in Dynamic Environments

Source: [DyPho-SLAM : Real-time Photorealistic SLAM in Dynamic Environments](<https://www.sota2.com/research/paper/68fa7b0e1e209131cea866cc>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-26T05:32:44.589Z

Authors: Yi Liu, Keyu Fan, Bin Lan, Houde Liu

Published: 2025-08-31T08:22:29.000Z

## Abstract

Visual SLAM algorithms have been enhanced through the exploration of Gaussian Splatting representations, particularly in generating high-fidelity dense maps. While existing methods perform reliably in static environments, they often encounter camera tracking drift and fuzzy mapping when dealing with the disturbances caused by moving objects. This paper presents DyPho-SLAM, a real-time, resource-efficient visual SLAM system designed to address the challenges of localization and photorealistic mapping in environments with dynamic objects. Specifically, the proposed system integrates prior image information to generate refined masks, effectively minimizing noise from mask misjudgment. Additionally, to enhance constraints for optimization after removing dynamic obstacles, we devise adaptive feature extraction strategies significantly improving the system's resilience. Experiments conducted on publicly dynamic RGB-D datasets demonstrate that the proposed system achieves state-of-the-art performance in camera pose estimation and dense map reconstruction, while operating in real-time in dynamic scenes.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Camera Tracking on TUM fr3 s half](<https://www.sota2.com/research/sota/camera-tracking-on-tum-fr3-s-half.md>): Camera Tracking; TUM fr3 s half; ATE (cm): 1.6.

- [Camera Tracking on Bonn crowd 2](<https://www.sota2.com/research/sota/camera-tracking-on-bonn-crowd-2.md>): Camera Tracking; Bonn crowd 2; ATE (cm): 2.5.

- [Camera Tracking on Bonn ps\_track](<https://www.sota2.com/research/sota/camera-tracking-on-bonn-ps_track.md>): Camera Tracking; Bonn ps\_track; ATE (cm): 3.7.

- [Camera Tracking on Bonn balloon 2](<https://www.sota2.com/research/sota/camera-tracking-on-bonn-balloon-2.md>): Camera Tracking; Bonn balloon 2; ATE (cm): 2.7.

- [Camera Tracking on TUM fr3 w xyz](<https://www.sota2.com/research/sota/camera-tracking-on-tum-fr3-w-xyz.md>): Camera Tracking; TUM fr3 w xyz; ATE (cm): 1.6.

- [Camera Tracking on TUM fr3 w half](<https://www.sota2.com/research/sota/camera-tracking-on-tum-fr3-w-half.md>): Camera Tracking; TUM fr3 w half; ATE (cm): 2.6.

- [Camera Tracking on Bonn balloon](<https://www.sota2.com/research/sota/camera-tracking-on-bonn-balloon.md>): Camera Tracking; Bonn balloon; Absolute Trajectory Error (cm): 3.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2509.00741>)

[PDF](<https://arxiv.org/pdf/2509.00741.pdf>)
