Unsupervised Domain Adaptation in LiDAR Semantic Segmentation with Self-Supervision and Gated Adapters

TL;DR

This paper presents an unsupervised domain adaptation framework for LiDAR semantic segmentation that uses self-supervision, mask transfer, and gated adapters to improve performance across different sensors and data sources.

Contribution

It introduces a novel combination of self-supervision, unpaired mask transfer, and gated adapters for effective domain adaptation in LiDAR segmentation.

Findings

Significant performance improvements over prior methods.

Effective adaptation from synthetic to real LiDAR data.

Robustness across different sensor domains.

Abstract

In this paper, we focus on a less explored, but more realistic and complex problem of domain adaptation in LiDAR semantic segmentation. There is a significant drop in performance of an existing segmentation model when training (source domain) and testing (target domain) data originate from different LiDAR sensors. To overcome this shortcoming, we propose an unsupervised domain adaptation framework that leverages unlabeled target domain data for self-supervision, coupled with an unpaired mask transfer strategy to mitigate the impact of domain shifts. Furthermore, we introduce the gated adapter module with a small number of parameters into the network to account for target domain-specific information. Experiments adapting from both real-to-real and synthetic-to-real LiDAR semantic segmentation benchmarks demonstrate the significant improvement over prior arts.