Bridging the Sim2Real gap with CARE: Supervised Detection Adaptation with Conditional Alignment and Reweighting
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.
Submission Length: Regular submission (no more than 12 pages of main content)
Changes Since Last Submission: Based on reviewer and AE feedback, we have made the following revisions:
* Related work (Sec 2): Added additional references requested by reviewers
* Approach (Sec 3.3.1): Simplified the notation for clarity
* Datasets and metrics (Sec 4.1): Included missing experimental details
* Additional backbones (Section 4.4): Included results with CARE on additional backbones
* Additional analysis (Appendix A.3): Added a complexity analysis for CARE
* Additional implementation details (Sec A.4): Added implementation details pertaining to training recipes and hyperparameter tuning
* Corrected typos.
Assigned Action Editor: ~Hongsheng_Li3
License: Creative Commons Attribution 4.0 International (CC BY 4.0)
Submission Number: 1341
Loading