CMDA: Cross-Modal and Domain Adversarial Adaptation for LiDAR-Based 3D Object Detection

TL;DR

This paper introduces CMDA, an unsupervised domain adaptation method for LiDAR-based 3D object detection that leverages visual cues and adversarial training to improve generalization across different data domains, achieving state-of-the-art results.

Contribution

The paper proposes a novel UDA framework, CMDA, combining cross-modal semantic bridging and self-training adversarial learning for better domain-invariant 3D detection.

Findings

Significant performance improvements on nuScenes, Waymo, and KITTI datasets.

Achieved state-of-the-art results in unsupervised domain adaptation for 3D object detection.

Effective use of visual semantic cues to bridge domain gaps.

Abstract

Recent LiDAR-based 3D Object Detection (3DOD) methods show promising results, but they often do not generalize well to target domains outside the source (or training) data distribution. To reduce such domain gaps and thus to make 3DOD models more generalizable, we introduce a novel unsupervised domain adaptation (UDA) method, called CMDA, which (i) leverages visual semantic cues from an image modality (i.e., camera images) as an effective semantic bridge to close the domain gap in the cross-modal Bird's Eye View (BEV) representations. Further, (ii) we also introduce a self-training-based learning strategy, wherein a model is adversarially trained to generate domain-invariant features, which disrupt the discrimination of whether a feature instance comes from a source or an unseen target domain. Overall, our CMDA framework guides the 3DOD model to generate highly informative and domain-adaptive features for novel data distributions. In our extensive experiments with large-scale benchmarks, such as nuScenes, Waymo, and KITTI, those mentioned above provide significant performance gains for UDA tasks, achieving state-of-the-art performance.