Spectral characteristics of Korean consonants based on a spontaneous speech corpus
0
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
The purpose of this study was to identify the spectral characteristics and patterns of Korean consonants. Consonant-vowel (CV) sequences were selected from the Seoul Corpus, a dialogue speech corpus. Fast Fourier transforms (FFTs) were applied, and the first four spectral moments were calculated: M1 (moment 1; center of gravity), M2 (moment 2; standard deviation), M3 (moment 3; skewness), and M4 (moment 4; kurtosis). Additionally, two slope-related parameters were measured: slope_ltas, which represents the dB difference between two frequency bands, and tilt, which indicates the regression slope. Six linear mixed-effects (LME) models were constructed to examine the effects of gender, manner of articulation, and vowel context on the dependent variables. The relationships between the acoustic variables were also investigated. Finally, binary classification accuracy for distinguishing obstruents from sonorants was calculated. Results indicated that gender differences appeared in obstruents and nasals using parameters other than M4. In distinguishing manners of articulation, the patterns observed in M1, slope_ltas, and tilt differed from those in M3 and M4. M2 showed a distinct trend compared to the other parameters. In terms of vowel context, M1 and M3 exhibited opposite patterns in their values. Furthermore, negative correlations were observed between M1 and M3, as well as between M3 and tilt. Lastly, the overall classification accuracy in distinguishing obstruents and sonorants was 67.67%. The most influential parameters were M3 and tilt.
Abstract
The purpose of this study was to identify the spectral characteristics and patterns of Korean consonants. Consonant-vowel (CV) sequences were selected from the Seoul Corpus, a dialogue speech corpus. Fast Fourier transforms (FFTs) were applied, and the first four spectral moments were calculated: M1 (moment 1; center of gravity), M2 (moment 2; standard deviation), M3 (moment 3; skewness), and M4 (moment 4; kurtosis). Additionally, two slope-related parameters were measured: slope_ltas, which represents the dB difference between two frequency bands, and tilt, which indicates the regression slope. Six linear mixed-effects (LME) models were constructed to examine the effects of gender, manner of articulation, and vowel context on the dependent variables. The relationships between the acoustic variables were also investigated. Finally, binary classification accuracy for distinguishing obstruents from sonorants was calculated. Results indicated that gender differences appeared in obstruents and nasals using parameters other than M4. In distinguishing manners of articulation, the patterns observed in M1, slope_ltas, and tilt differed from those in M3 and M4. M2 showed a distinct trend compared to the other parameters. In terms of vowel context, M1 and M3 exhibited opposite patterns in their values. Furthermore, negative correlations were observed between M1 and M3, as well as between M3 and tilt. Lastly, the overall classification accuracy in distinguishing obstruents and sonorants was 67.67%. The most influential parameters were M3 and tilt.
참고문헌 (23)
- (2012)[URL]
- (2020)[URL]
- Flipsen P. Jr, "Acoustic characteristics of", s, Journal of Speech, Language, and Hearing Research, pp.1999, 42(3)[DOI]
- Forrest, K., "Statistical analysis of word-initial voiceless obstruents: Preliminary data", The Journal of the Acoustical Society of America, 84(1), pp.115~123 (1988)
- Han, K. I., "Temporal and spectral characteristics at the midpoint of frication noise in Korean fricatives", Journal of Language Sciences, 23(3), pp.277~300 (2016)
- Heo, H. J., "Acoustic characteristics of the Korean alveolar fricative", s, EONEOHAG, pp.2015, (72)[DOI]
- Hwang, H. K., "Spectral characteristics of frication noise in Korean sibilants", MALSORI, (49), pp.31~50 (2004)
- Johnson, K., "Acoustic and auditory phonetics" (2011), Wiley-Blackwell
- Jongman, A., "Phonetics of fricatives", Oxford Research Encyclopedia of Linguistics (2024)[DOI]
- Jongman, A., "Acoustic characteristics of English fricatives", The Journal of the Acoustical Society of America, 108(3), pp.1252~1263 (2000)
- Kim, J., "A perceptual and acoustical study of", s, Phonetics and Speech Sciences, pp.2018, 10(3)[DOI]
- Nissen, S. L., "Acoustic and spectral characteristics of young children's fricative productions: A developmental perspective", The Journal of the Acoustical Society of America, 118(4), pp.2570~2578 (2005)
- Park, H., "Spectral characteristics of release bursts", Proceedings of the KSPS Conference, pp.159~162 (2003)
- Park, S., "Analysis and selection of acoustic parameters for classifying places of articulation and phonation types of Korean plosive sounds", HAN-GEUL, 83(3), pp.571~610 (2022)
- (2024)[URL]
- Tabain, M., "An acoustic study of multiple lateral consonants in three Central Australian languages", The Journal of the Acoustical Society of America, 139(1), pp.361~372 (2016)
- Tabain, M., "An acoustic study of nasal consonants in three Central Australian languages", The Journal of the Acoustical Society of America, 139(2), pp.890~903 (2016)
- Tabain, M., "Acoustic realization and inventory size: Kannada and Malayalam alveolar", retroflex laterals and, Phonetica, pp.2018, 75(2)[DOI]
- Taylor, S., "Age-related changes in speech and voice: Spectral and cepstral measures", Journal of Speech, Language, and Hearing Research, 63(3), pp.647~660 (2020)
- Themistocleous, C., "Dialect classification from a single sonorant sound using deep neural networks", Frontiers in Communication, 4, pp.64~ (2019)
- Themistocleous, C., "Sonorant spectra and coarticulation distinguish speakers with different dialects", Speech Communication, 142, pp.1~14 (2022)
- Yoon, K., "A study on the Korean plosives from the Seoul corpus of spontaneous speech", Language Facts and Perspectives, 50, pp.479~501 (2020)
- Yun, W., "The Korean corpus of spontaneous speech", Journal of the Korean Society of Speech Sciences, 7(2), pp.103~109 (2015)