Evaluating the Impact of Semantic Segmentation and Pose Estimation on Dense Semantic SLAM

TL;DR

This paper evaluates how semantic segmentation and pose estimation affect the quality of dense semantic SLAM maps, revealing that segmentation errors significantly degrade performance.

Contribution

It provides a comprehensive evaluation of semantic SLAM algorithms using ground-truth data, highlighting the impact of segmentation and pose estimation errors on map quality.

Findings

Semantic segmentation errors cause up to 74.3% drop in mAP.

Pose estimation errors have a lesser impact compared to segmentation.

Ground-truth data helps isolate sources of errors in semantic SLAM.

Abstract

Recent Semantic SLAM methods combine classical geometry-based estimation with deep learning-based object detection or semantic segmentation. In this paper we evaluate the quality of semantic maps generated by state-of-the-art class- and instance-aware dense semantic SLAM algorithms whose codes are publicly available and explore the impacts both semantic segmentation and pose estimation have on the quality of semantic maps. We obtain these results by providing algorithms with ground-truth pose and/or semantic segmentation data available from simulated environments. We establish that semantic segmentation is the largest source of error through our experiments, dropping mAP and OMQ performance by up to 74.3% and 71.3% respectively.