JAIST Repository >
School of Information Science >
Conference Papers >
Conference Papers >

Please use this identifier to cite or link to this item: http://hdl.handle.net/10119/15775

Title: Study on Nonlinear Relationships between Semantic Primitives and Emotional Dimensions for Improving Three-layered Model
Authors: Liu, Xingyu
Elbarougy, Reda Elsaid
Akagi, Masato
Issue Date: 2019-03-07
Publisher: Research Institute of Signal Processing, Japan
Magazine name: 2019 RISP International Workshop on Nonlinear Circuits, Communications and Signal Processing (NCSP2019)
Start page: 522
End page: 525
Abstract: Three-layered model is a perceptual model mimicking the process of human perception on speech emotion. However, previous studies based on the three-layered model only focused on the linear relationships in the emotion perception process among the three layers. In this research, nonlinear relationships between the second layer (semantic primitives) and the first layer (emotion dimensions), the top two layers of the three-layered model, were investigated by taking advantage of fuzzy inference system (FIS) as an estimator. Effective semantic primitives to describe human perception on emotion dimensions were selected from 28 semantic primitive candidates. Evaluation results show that semantic primitives selected by the proposed method are effective to describe emotion dimensions and can be used to construct an improved three-layered model.
Rights: Copyright (C) 2019 Research Institute of Signal Processing, Japan. Xingyu Liu, Reda Elsaid Elbarougy, and Masato Akagi, 2019 RISP International Workshop on Nonlinear Circuits, Communications and Signal Processing (NCSP2019), 2019, 522-525.
URI: http://hdl.handle.net/10119/15775
Material Type: publisher
Appears in Collections:b11-1. 会議発表論文・発表資料 (Conference Papers)

Files in This Item:

File Description SizeFormat
2924.pdf730KbAdobe PDFView/Open

All items in DSpace are protected by copyright, with all rights reserved.

 


Contact : Library Information Section, JAIST (ir-sys[at]ml.jaist.ac.jp)