Deep Adaptive Attention for Joint Facial Action Unit Detection and Face Alignment

TL;DR

This paper introduces a novel end-to-end deep learning framework that jointly performs facial action unit detection and face alignment, leveraging adaptive attention and shared multi-scale features to improve accuracy.

Contribution

It proposes the first integrated deep learning model for simultaneous AU detection and face alignment with adaptive attention mechanisms.

Findings

Outperforms state-of-the-art AU detection methods on BP4D and DISFA datasets.

Effectively combines face alignment features with local and global features for improved detection.

Demonstrates the benefit of joint learning over separate task handling.

Abstract

Facial action unit (AU) detection and face alignment are two highly correlated tasks since facial landmarks can provide precise AU locations to facilitate the extraction of meaningful local features for AU detection. Most existing AU detection works often treat face alignment as a preprocessing and handle the two tasks independently. In this paper, we propose a novel end-to-end deep learning framework for joint AU detection and face alignment, which has not been explored before. In particular, multi-scale shared features are learned firstly, and high-level features of face alignment are fed into AU detection. Moreover, to extract precise local features, we propose an adaptive attention learning module to refine the attention map of each AU adaptively. Finally, the assembled local features are integrated with face alignment features and global features for AU detection. Experiments on BP4D and DISFA benchmarks demonstrate that our framework significantly outperforms the state-of-the-art methods for AU detection.