Deep Region and Multi-label Learning for Facial Action Unit Detection
Authors:
- Kaili ZhaoKZKaili Zhao
- Wen-Sheng ChuWCWen-Sheng Chu
- Honggang ZhangHZHonggang Zhang
- 1 June 2016
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 3391-3399
Abstract
Region learning (RL) and multi-label learning (ML) have recently attracted increasing attentions in the field of facial Action Unit (AU) detection. Knowing that AUs are active on sparse facial regions, RL aims to identify these regions for a better specificity. On the other hand, a strong statistical evidence of AU correlations suggests that ML is a natural way to model the detection task. In this paper, we propose Deep Region and Multi-label Learning (DRML), a unified deep network that simultaneously addresses these two problems. One crucial aspect in DRML is a novel region layer that uses feed-forward functions to induce important facial regions, forcing the learned weights to capture structural information of the face. Our region layer serves as an alternative design between locally connected layers (i.e., confined kernels to individual pixels) and conventional convolution layers (i.e., shared kernels across an entire image). Unlike previous studies that solve RL and ML alternately, DRML by construction addresses both problems, allowing the two seemingly irrelevant problems to interact more directly. The complete network is end-to-end trainable, and automatically learns representations robust to variations inherent within a local region. Experiments on BP4D and DISFA benchmarks show that DRML performs the highest average F1-score and AUC within and across datasets in comparison with alternative methods.Keywords
This publication has 20 references indexed in Scilit:
- Confidence Preserving Machine for Facial Action Unit Detection2015
- Joint patch and multi-label learning for facial action unit detection2015
- Caffe2014
- DeepFace: Closing the Gap to Human-Level Performance in Face Verification2014
- Compound facial expressions of emotionProceedings of the National Academy of Sciences, 2014
- A high-resolution spontaneous 3D dynamic facial expression database2013
- Learning hierarchical representations for face verification with convolutional deep belief networks2012
- Action unit detection using sparse appearance descriptors in space-time video volumes2011
- Investigating Spontaneous Facial Action Recognition through AAM Representations of the Face2007
- Dynamics of facial expression extracted automatically from videoImage and Vision Computing, 2006