MobileFAN: Transferring deep hidden representation for face alignment. (April 2020)

• Record Type: Journal Article
• Authors: Zhao, Yang; Liu, Yifan; Shen, Chunhua; Gao, Yongsheng; Xiong, Shengwu
• Is Part Of: Pattern recognition. Volume 100(2020:Apr.)
• Publication Date: 2020-04
• Subjects: Face alignment -- Knowledge distillation -- Lightweight model; Pattern perception -- Periodicals; Perception des structures -- Périodiques; Patroonherkenning; 006.4
• DOI: 10.1016/j.patcog.2019.107114
• Languages: English
• ISSNs: 0031-3203

Abstract

Highlights: A simple and lightweight network, namely Mobile Face Alignment Network (MobileFAN), is proposed for facial landmark detection. It can effectively handle face alignment problem in high accuracy with only 8% of the model size of state-of-the-art models. This is the first attempt to introduce knowledge distillation techniques to heatmap regression-based method for performance enhancement in face alignment. The experimental results on three public available datasets demonstrate the effectiveness and superiority of the proposed model. Abstract: Facial landmark detection is a crucial prerequisite for many face analysis applications. Deep learning-based methods currently dominate the approach of addressing the facial landmark detection. However, such works generally introduce a large number of parameters, resulting in high memory cost. In this paper, we aim for a lightweight as well as effective solution to facial landmark detection. To this end, we propose an effective lightweight model, namely Mobile Face Alignment Network (MobileFAN), using a simple backbone MobileNetV2 as the encoder and three deconvolutional layers as the decoder. The proposed MobileFAN, with only 8% of the model size and lower computational cost, achieves superior or equivalent performance compared with state-of-the-art models. Moreover, by transferring the geometric structural information of a face graph from a large complex model to our proposed MobileFAN through feature-aligned distillation and feature-similarity distillation, the performance of MobileFAN is further improved in effectiveness and efficiency for face alignment. Extensive experiment results on three challenging facial landmark estimation benchmarks including COFW, 300W and WFLW show the superiority of our proposed MobileFAN against state-of-the-art methods.