2024-03-29T01:54:32Z
https://meral.edu.mm/oai
oai:meral.edu.mm:recid/4999
2021-12-13T02:42:51Z
1582963302567:1597824273898
user-ucsy
Cross-modal Sentiment Information Expression of Voice Source Characteristics using Image Texture Features
Kyaw, Win Thuzar
SAGISAKA, Yoshinori
Following the successful findings of highcorrelations between speech and color such as F0 andValue, Loudness and Saturation and Spectrum andHue, we analyzed the correlations between voicesource characteristics and the image parametersshowing textural differences in this paper for betterscientific understanding of their correlations andeffective use in visualization of speech information.Through sentiment association experiments, we couldhave observed high positive correlations between H1*-H2* (amplitude difference between first and secondharmonics corrected for vocal tract effects), H1-A1(amplitude difference between first harmonic and firstformant) and Contrast, high negative correlationsbetween H1*-H2*, H1-A1, H1-A2, H1-A3, Harmonicto-Noise Ratio (HNR) in 0 to 3500Hz frequency bandand Variance, Prominence and negative correlationsbetween H1*-A3*, HNR in 0 to 500 Hz andProminence. These results show the possibility ofdirect visualization of speech characteristics whichcannot be effectively carried out by conventionalmapping using discrete language expressions.
2017-02-16
http://hdl.handle.net/20.500.12678/0000004999
https://meral.edu.mm/records/4999