Bridging the Sim2Real gap with CARE: Supervised Detection Adaptation with Conditional Alignment and Reweighting

TL;DR

This paper introduces CARE, a novel supervised domain adaptation method for 2D object detection that leverages limited real labels to effectively bridge the sim2real gap by aligning appearance and content.

Contribution

CARE systematically exploits target labels to explicitly close appearance and content gaps in supervised sim2real domain adaptation for object detection.

Findings

CARE outperforms competing methods on standard benchmarks.

Explicit alignment improves detection accuracy in sim2real tasks.

Analytical justification supports the effectiveness of CARE.

Abstract

Sim2Real domain adaptation (DA) research focuses on the constrained setting of adapting from a labeled synthetic source domain to an unlabeled or sparsely labeled real target domain. However, for high-stakes applications (e.g. autonomous driving), it is common to have a modest amount of human-labeled real data in addition to plentiful auto-labeled source data (e.g. from a driving simulator). We study this setting of supervised sim2real DA applied to 2D object detection. We propose Domain Translation via Conditional Alignment and Reweighting (CARE) a novel algorithm that systematically exploits target labels to explicitly close the sim2real appearance and content gaps. We present an analytical justification of our algorithm and demonstrate strong gains over competing methods on standard benchmarks.