Denoising and Selecting Pseudo-Heatmaps for Semi-Supervised Human Pose Estimation

TL;DR

This paper introduces a semi-supervised human pose estimation method that denoises pseudo-heatmaps and selects targets based on uncertainty, significantly improving performance especially with limited labeled data.

Contribution

It presents a novel denoising scheme and uncertainty-guided target selection within a dual-student framework for semi-supervised pose estimation.

Findings

Outperforms previous methods on COCO benchmark.

Surpasses state-of-the-art with only 0.5K labeled images.

Effectively leverages unlabeled data for better generalization.

Abstract

We propose a new semi-supervised learning design for human pose estimation that revisits the popular dual-student framework and enhances it two ways. First, we introduce a denoising scheme to generate reliable pseudo-heatmaps as targets for learning from unlabeled data. This uses multi-view augmentations and a threshold-and-refine procedure to produce a pool of pseudo-heatmaps. Second, we select the learning targets from these pseudo-heatmaps guided by the estimated cross-student uncertainty. We evaluate our proposed method on multiple evaluation setups on the COCO benchmark. Our results show that our model outperforms previous state-of-the-art semi-supervised pose estimators, especially in extreme low-data regime. For example with only 0.5K labeled images our method is capable of surpassing the best competitor by 7.22 mAP (+25% absolute improvement). We also demonstrate that our model can learn effectively from unlabeled data in the wild to further boost its generalization and performance.