MAPS: A Noise-Robust Progressive Learning Approach for Source-Free Domain Adaptive Keypoint Detection

TL;DR

This paper introduces MAPS, a novel source-free domain adaptation method for keypoint detection that leverages progressive pseudo-label selection and mixup augmentation to improve noise robustness and performance.

Contribution

The paper proposes MAPS, a unified approach combining self-mixup augmentation and self-paced learning for source-free domain adaptive keypoint detection, addressing privacy and security concerns.

Findings

MAPS outperforms baseline methods on four datasets.

MAPS achieves comparable or better results than non-source-free methods.

The approach effectively handles noisy pseudo labels during training.

Abstract

Existing cross-domain keypoint detection methods always require accessing the source data during adaptation, which may violate the data privacy law and pose serious security concerns. Instead, this paper considers a realistic problem setting called source-free domain adaptive keypoint detection, where only the well-trained source model is provided to the target domain. For the challenging problem, we first construct a teacher-student learning baseline by stabilizing the predictions under data augmentation and network ensembles. Built on this, we further propose a unified approach, Mixup Augmentation and Progressive Selection (MAPS), to fully exploit the noisy pseudo labels of unlabeled target data during training. On the one hand, MAPS regularizes the model to favor simple linear behavior in-between the target samples via self-mixup augmentation, preventing the model from over-fitting to noisy predictions. On the other hand, MAPS employs the self-paced learning paradigm and progressively selects pseudo-labeled samples from `easy' to `hard' into the training process to reduce noise accumulation. Results on four keypoint detection datasets show that MAPS outperforms the baseline and achieves comparable or even better results in comparison to previous non-source-free counterparts.