Robust Face Alignment by Multi-order High-precision Hourglass Network

TL;DR

This paper introduces a novel multi-order high-precision hourglass network with heatmap subpixel regression and cross geometry-aware modeling, significantly improving face alignment accuracy under challenging conditions like large poses and occlusions.

Contribution

It presents the first heatmap subpixel regression technique and integrates it with a multi-order cross geometry-aware model into a high-precision hourglass network for robust face alignment.

Findings

Outperforms state-of-the-art methods on benchmark datasets.

Achieves high-precision landmark detection under challenging conditions.

Effectively handles large pose variations and occlusions.

Abstract

Heatmap regression (HR) has become one of the mainstream approaches for face alignment and has obtained promising results under constrained environments. However, when a face image suffers from large pose variations, heavy occlusions and complicated illuminations, the performances of HR methods degrade greatly due to the low resolutions of the generated landmark heatmaps and the exclusion of important high-order information that can be used to learn more discriminative features. To address the alignment problem for faces with extremely large poses and heavy occlusions, this paper proposes a heatmap subpixel regression (HSR) method and a multi-order cross geometry-aware (MCG) model, which are seamlessly integrated into a novel multi-order high-precision hourglass network (MHHN). The HSR method is proposed to achieve high-precision landmark detection by a well-designed subpixel detection loss (SDL) and subpixel detection technology (SDT). At the same time, the MCG model is able to use the proposed multi-order cross information to learn more discriminative representations for enhancing facial geometric constraints and context information. To the best of our knowledge, this is the first study to explore heatmap subpixel regression for robust and high-precision face alignment. The experimental results from challenging benchmark datasets demonstrate that our approach outperforms state-of-the-art methods in the literature.