Laughter classification using 3D convolutional neural networks
College
College of Computer Studies
Department/Unit
Software Technology
Document Type
Conference Proceeding
Source Title
ACM International Conference Proceeding Series
First Page
47
Last Page
51
Publication Date
10-26-2019
Abstract
Social signals express the attitude of human being in social situations. Laughter has been determined as an important social signal that can predict emotional information of people. It conveys different emotions such as happiness, surprise, fear, anger, and anxiety. Therefore, identifying and extracting emotions in the laughter is useful for estimating the emotional state of the user. Deep neural networks are replacing traditional methods because they perform more accurately. This paper presents work that detects the emotions in laughter by using audio features and running 3D Convolutional Neural Networks. The best rate of accuracy produced by 3D CNNs is 97.97%, which is higher than the results of our previous paper, which applied MLP and SVM on Iranian laughter dataset. © 2019 ACM.
html
Digitial Object Identifier (DOI)
10.1145/3369114.3369142
Recommended Citation
Ataollahi, F., & Suarez, M. (2019). Laughter classification using 3D convolutional neural networks. ACM International Conference Proceeding Series, 47-51. https://doi.org/10.1145/3369114.3369142
Disciplines
Computer Sciences | Software Engineering
Keywords
Emotion recognition; Laughter; Neural networks (Computer science)
Upload File
wf_no