Simulation-to-Reality domain adaptation for offline 3D object annotation on pointclouds with correlation alignment

TL;DR

This paper introduces a semi-automatic method for annotating real-world LiDAR pointclouds by leveraging simulated data and correlation alignment to reduce domain gap in 3D object detection.

Contribution

It proposes a novel domain adaptation approach using CORAL loss to align simulated and real pointcloud features for improved 3D object annotation.

Findings

Enhanced annotation efficiency for real-world pointclouds.

Effective domain invariance achieved through correlation alignment.

Improved 3D object detection performance in real data.

Abstract

Annotating objects with 3D bounding boxes in LiDAR pointclouds is a costly human driven process in an autonomous driving perception system. In this paper, we present a method to semi-automatically annotate real-world pointclouds collected by deployment vehicles using simulated data. We train a 3D object detector model on labeled simulated data from CARLA jointly with real world pointclouds from our target vehicle. The supervised object detection loss is augmented with a CORAL loss term to reduce the distance between labeled simulated and unlabeled real pointcloud feature representations. The goal here is to learn representations that are invariant to simulated (labeled) and real-world (unlabeled) target domains. We also provide an updated survey on domain adaptation methods for pointclouds.