Cross-Task Inconsistency Based Active Learning (CTIAL) for Emotion Recognition

被引:1
作者
Xu, Yifan [1 ]
Jiang, Xue [1 ]
Wu, Dongrui [1 ]
机构
[1] Huazhong Univ Sci & Technol, Sch Artificial Intelligence & Automat, Key Lab, Minist Educ Image Proc & Intelligent Control, Wuhan 430074, Peoples R China
关键词
Task analysis; Uncertainty; Measurement uncertainty; Estimation; Entropy; Emotion recognition; Affective computing; Active learning; emotion classification; emotion estimation; transfer learning; FACIAL EXPRESSIONS;
D O I
10.1109/TAFFC.2024.3366767
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Emotion recognition is a critical component of affective computing. Training accurate machine learning models for emotion recognition typically requires a large amount of labeled data. Due to the subtleness and complexity of emotions, multiple evaluators are usually needed for each affective sample to obtain its ground-truth label, which is expensive. To save the labeling cost, this paper proposes an inconsistency-based active learning approach for cross-task transfer between emotion classification and estimation. Affective norms are utilized as prior knowledge to connect the label spaces of categorical and dimensional emotions. Then, the prediction inconsistency on the two tasks for the unlabeled samples is used to guide sample selection in active learning for the target task. Experiments on within-corpus and cross-corpus transfers demonstrated that cross-task inconsistency could be a very valuable metric in active learning. To our knowledge, this is the first work that utilizes prior knowledge on affective norms and data in a different task to facilitate active learning for a new task, even the two tasks are from different datasets.
引用
收藏
页码:1659 / 1668
页数:10
相关论文
共 40 条
[1]  
Abdelwahab M, 2019, INT CONF AFFECT, DOI [10.1109/ACII.2019.8925524, 10.1109/acii.2019.8925524]
[2]   Emotion Based Music Recommendation System Using Wearable Physiological Sensors [J].
Ayata, Deger ;
Yaslan, Yusuf ;
Kamasak, Mustafa E. .
IEEE TRANSACTIONS ON CONSUMER ELECTRONICS, 2018, 64 (02) :196-203
[3]  
Baevski A, 2020, ADV NEUR IN, V33
[4]   Integrating structured biological data by Kernel Maximum Mean Discrepancy [J].
Borgwardt, Karsten M. ;
Gretton, Arthur ;
Rasch, Malte J. ;
Kriegel, Hans-Peter ;
Schoelkopf, Bernhard ;
Smola, Alex J. .
BIOINFORMATICS, 2006, 22 (14) :E49-E57
[5]  
Bradley MargaretM., 1999, AFFECTIVE NORMS ENGL
[6]   IEMOCAP: interactive emotional dyadic motion capture database [J].
Busso, Carlos ;
Bulut, Murtaza ;
Lee, Chi-Chun ;
Kazemzadeh, Abe ;
Mower, Emily ;
Kim, Samuel ;
Chang, Jeannette N. ;
Lee, Sungbok ;
Narayanan, Shrikanth S. .
LANGUAGE RESOURCES AND EVALUATION, 2008, 42 (04) :335-359
[7]  
Culotta A., 2005, Proceedings of the 20th National Conference on Artificial Intelligence (AAAI), P746
[8]   UNIVERSALS AND CULTURAL-DIFFERENCES IN THE JUDGMENTS OF FACIAL EXPRESSIONS OF EMOTION [J].
EKMAN, P ;
FRIESEN, WV ;
OSULLIVAN, M ;
CHAN, A ;
DIACOYANNITARLATZIS, I ;
HEIDER, K ;
KRAUSE, R ;
LECOMPTE, WA ;
PITCAIRN, T ;
RICCIBITTI, PE ;
SCHERER, K ;
TOMITA, M ;
TZAVARAS, A .
JOURNAL OF PERSONALITY AND SOCIAL PSYCHOLOGY, 1987, 53 (04) :712-717
[9]  
El Kaliouby R., 2004, 2004 Conference on Computer Vision and Pattern Recognition Workshop, P154, DOI DOI 10.1109/CVPR.2004.153
[10]   THE VERA AM MITTAG GERMAN AUDIO-VISUAL EMOTIONAL SPEECH DATABASE [J].
Grimm, Michael ;
Kroschel, Kristian ;
Narayanan, Shrikanth .
2008 IEEE INTERNATIONAL CONFERENCE ON MULTIMEDIA AND EXPO, VOLS 1-4, 2008, :865-+