Meta Auxiliary Learning for Facial Action Unit Detection

被引:7
|
作者
Li, Yong [1 ,2 ]
Shan, Shiguang [3 ,4 ]
机构
[1] Nanjing Univ Sci & Technol, PCA Lab, Minist Educ, Key Lab Intelligent Percept & Syst High Dimens inf, Nanjing 210094, Peoples R China
[2] Nanjing Univ Sci & Technol, Sch Comp Sci & Engn, Jiangsu Key Lab Image & Video Understanding Social, Nanjing 210094, Peoples R China
[3] Chinese Acad Sci, Key Lab Intelligent Informat Proc, Inst Comp Technol, Beijing 100190, Peoples R China
[4] CAS Ctr Excellencein Brain Sci & Intelligence Tech, Shanghai 200031, Peoples R China
基金
国家重点研发计划; 中国国家自然科学基金;
关键词
Facial action unit detection; auxiliary learning; meta learning; EXPRESSIONS;
D O I
10.1109/TAFFC.2021.3135516
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Despite the success of deep neural networks on facial action unit (AU) detection, better performance depends on a large number of training images with accurate AU annotations. However, labeling AU is time-consuming, expensive, and error-prone. Considering AU detection and facial expression recognition (FER) are two highly correlated tasks, and facial expression (FE) is relatively easy to annotate, we consider learning AU detection and FER in a multi-task manner. However, the performance of the AU detection task cannot be always enhanced due to the negative transfer in the multi-task scenario. To alleviate this issue, we propose a Meta Auxiliary Learning method (MAL) that automatically selects highly related FE samples by learning adaptative weights for the training FE samples in a meta learning manner. The learned sample weights alleviate the negative transfer from two aspects: 1) balance the loss of each task automatically, and 2) suppress the weights of FE samples that have large uncertainties. Experimental results on several popular AU datasets demonstrate MAL consistently improves the AU detection performance compared with the state-of-the-art multi-task and auxiliary learning methods. MAL automatically estimates adaptive weights for the auxiliary FE samples according to their semantic relevance with the primary AU detection task.
引用
收藏
页码:2526 / 2538
页数:13
相关论文
共 50 条
  • [21] Learning Spatial and Temporal Cues for Multi-label Facial Action Unit Detection
    Chu, Wen-Sheng
    De la Torre, Fernando
    Cohn, Jeffrey F.
    [J]. 2017 12TH IEEE INTERNATIONAL CONFERENCE ON AUTOMATIC FACE AND GESTURE RECOGNITION (FG 2017), 2017, : 25 - 32
  • [22] Contrastive Feature Learning and Class-Weighted Loss for Facial Action Unit Detection
    Wu, Bing-Fei
    Wei, Yin-Tse
    Wu, Bing-Jhang
    Lin, Chun-Hsien
    [J]. 2019 IEEE INTERNATIONAL CONFERENCE ON SYSTEMS, MAN AND CYBERNETICS (SMC), 2019, : 2478 - 2483
  • [23] Cross-Modal Representation Learning for Lightweight and Accurate Facial Action Unit Detection
    Chen, Yingjie
    Wu, Han
    Wang, Tao
    Wang, Yizhou
    Liang, Yun
    [J]. IEEE ROBOTICS AND AUTOMATION LETTERS, 2021, 6 (04): : 7619 - 7626
  • [24] Heterogeneous spatio-temporal relation learning network for facial action unit detection
    Song, Wenyu
    Shi, Shuze
    Dong, Yu
    An, Gaoyun
    [J]. PATTERN RECOGNITION LETTERS, 2022, 164 : 268 - 275
  • [25] Self-supervised Representation Learning from Videos for Facial Action Unit Detection
    Li, Yong
    Zeng, Jiabei
    Shan, Shiguang
    Chen, Xilin
    [J]. 2019 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR 2019), 2019, : 10916 - 10925
  • [26] Facial action unit detection with emotion consistency: a cross-modal learning approach
    Song, Wenyu
    Liu, Dongxin
    An, Gaoyun
    Duan, Yun
    Wang, Laifu
    [J]. Multimedia Systems, 2024, 30 (06)
  • [27] Multi-Scale Region with Local Relationship Learning for Facial Action Unit Detection
    Shi, Shuze
    An, Gaoyun
    Ruan, Qiuqi
    [J]. PROCEEDINGS OF 2020 IEEE 15TH INTERNATIONAL CONFERENCE ON SIGNAL PROCESSING (ICSP 2020), 2020, : 252 - 256
  • [28] Facial Action Unit detection based on multi-task learning strategy for unlabeled facial images in the wild
    Shang, Ziqiao
    Liu, Bin
    [J]. EXPERT SYSTEMS WITH APPLICATIONS, 2024, 253
  • [29] Facial Action Unit Detection Based on Teacher-Student Learning Framework for Partially Occluded Facial Images
    Kawamura, Ryosuke
    Murase, Kentaro
    [J]. 2021 16TH IEEE INTERNATIONAL CONFERENCE ON AUTOMATIC FACE AND GESTURE RECOGNITION (FG 2021), 2021,
  • [30] Stacking multiple cues for facial action unit detection
    Simge Akay
    Nafiz Arica
    [J]. The Visual Computer, 2022, 38 : 4235 - 4250