Low-resolution Human Pose Estimation

TL;DR

This paper introduces a Confidence-Aware Learning method to improve low-resolution human pose estimation by addressing limitations of existing offset learning techniques, resulting in significant performance gains on the COCO benchmark.

Contribution

The paper proposes a novel CAL approach that enhances offset learning for low-resolution images, improving robustness and accuracy over existing methods.

Findings

Outperforms state-of-the-art methods on COCO benchmark

Offset learning is more affected by low resolution, CAL mitigates this

CAL improves model consistency between training and testing phases

Abstract

Human pose estimation has achieved significant progress on images with high imaging resolution. However, low-resolution imagery data bring nontrivial challenges which are still under-studied. To fill this gap, we start with investigating existing methods and reveal that the most dominant heatmap-based methods would suffer more severe model performance degradation from low-resolution, and offset learning is an effective strategy. Established on this observation, in this work we propose a novel Confidence-Aware Learning (CAL) method which further addresses two fundamental limitations of existing offset learning methods: inconsistent training and testing, decoupled heatmap and offset learning. Specifically, CAL selectively weighs the learning of heatmap and offset with respect to ground-truth and most confident prediction, whilst capturing the statistical importance of model output in mini-batch learning manner. Extensive experiments conducted on the COCO benchmark show that our method outperforms significantly the state-of-the-art methods for low-resolution human pose estimation.