Unsupervised Domain Adaptation for 3D Keypoint Estimation via View Consistency

TL;DR

This paper proposes an unsupervised domain adaptation method for 3D keypoint estimation that leverages view consistency and geometric alignment to improve predictions on unlabeled data, outperforming existing techniques.

Contribution

It introduces a novel view consistency regularization and geometric alignment approach for unsupervised domain adaptation in 3D keypoint prediction.

Findings

Outperforms state-of-the-art domain adaptation methods.

Effective regularization via view consistency improves accuracy.

Demonstrates robustness on real datasets.

Abstract

In this paper, we introduce a novel unsupervised domain adaptation technique for the task of 3D keypoint prediction from a single depth scan or image. Our key idea is to utilize the fact that predictions from different views of the same or similar objects should be consistent with each other. Such view consistency can provide effective regularization for keypoint prediction on unlabeled instances. In addition, we introduce a geometric alignment term to regularize predictions in the target domain. The resulting loss function can be effectively optimized via alternating minimization. We demonstrate the effectiveness of our approach on real datasets and present experimental results showing that our approach is superior to state-of-the-art general-purpose domain adaptation techniques.