2008年6月11日 星期三

[音韻] Part Five: Ch 23.3~6 How do Listeners Compensate for Phonology?

Part Five: Ch 23.3~6 How do Listeners Compensate for Phonology?
23.3 Experiment 2
23.3.1 Methods
Use monomorphemic non-words in (C)VCCV(V)
A linguistically naïve female native Korean speaker recorded the stimuli
310 different items occurred twice
Ten Korean native speakers as listeners
23.3.2. Results
Listeners show high rates of correct response(96.8%) to the fillers.
23.3.3 Discussion
Goal
Whether listeners can infer the underlying forms
From phonetic homorganic clusters in non-words
Interesting result
Velar stimuli
Both /s.k/ and /p.k/ assimilate to [k.k] in natural speech
Main factors of lowering YES response
Frequency of labial-to-velar assimilation
Coronal-to-noncornoal assimilation
23.4 Experiment 3
23.4.1 Methods
Stimuli is pseudo-compounds consisting of two words
Listeners were provided transcription options
Ten Korean native speakers
23.4.2. Results
Rate of YES is 96.3%
Subject understood the task
23.4.3 Discussion
Purpose
Test the expectation
If listeners identified the initial morpheme in a new compound
Response
Higher than the previous ones
23.5 General Discussion
Subject are
preferred homorganic clusters as underlying forms
preferred coronal as targets of assimilation
Response pattern in Experiment 2 &3
Support the phonological-inference account
Listeners did not recover legal heterorganic clusters as underlying form
Result of Experiment 3
Show listeners reply more on phonological inference.
23.6 Conclusion
Goal
To test whether listeners use their phonological knowledge to infer original forms in speech recognition
Result suggest
Phonological inference is a part of the speech recognition process

[音韻] Max Planck Institute for Psycholinguistics

http://www.mpi.nl/

2008年6月8日 星期日

[音韻] Further reading of Ch16.3~4

1.《Acoustical society of America Digital Library》http://scitation.aip.org/getabs/servlet/GetabsServlet?prog=normal&id=JASMAN00012200000200EL35000001&idtype=cvips&gifs=yes
2. 〈HST 725 Lecture 6 Temporal models for musical pitch〉Howard-MIT Division if Health Science & Technology, HST 725: Music Perception & cognitive, Prof Peter Carinar

http://ocw.mit.edu/NR/rdonlyres/Health-Sciences-and-Technology/HST-725Spring2004/C75E3710-A429-491B-BA4B-3AAE5DC88338/0/lecture6.pdf
3. Peter Assmann
http://www.utdallas.edu/~assmann/
4. Terreance Nearey
http://www.ualberta.ca/~tnearey/
5. Ch11: Controls and Data entry devices〉輔大心理系
http://psy.vb-net.org/HF2007/2007ch11.ppt#257,1,Ch11. Controls and Data Entry Devices 控制器與資料輸入裝置
6. Ψ
http://en.wikipedia.org/wiki/Psi_(letter)
Ψ
The letter psi is commonly used in physics for representing a wavefunction in quantum mechanics, particularly with the Schrödinger equation and bra-ket notation: . It is also used to represent the (generalized) positional states of a qubit in a quantum computer.
Psi is also used as the symbol for the polygamma function, defined by
where Γ(x) is the gamma function.

[音韻] Part four: Probabilistic “Sliding template” models for indirect vowel normalization Ch 16.3~4

Part four: Probabilistic “Sliding template” models for indirect vowel normalization Ch 16.3~4
16.3.Normalization with a latent
16.3.1 Method 2: Unrestricted (or uniform prior) optimization over
Need the tokens from all of a speaker’s vowel categories to estimate
Several method
Related to methods of vocaltract length normalization
Spectral scaling
Spectral warping in the ASR literature(Westphal 1997)
The intermediate expression Pv(2) as an implicit shorthand for
Fuller expression
Arg max [P(Gt︱v, Ψs)]
Both Equation (11) and Equation (12)
Choose the vowel that looks best when it tries to look best
Equation (11)
Each vowel, the template us slide along a track on the table
Equation (12)
The vowel patch with the greatest light transmission
The scheme imposed no restrictions or preference onΨs
16.3.2 Method 3: Optimization of with an informative prior
Evidence
Chiba and kijiyama(1958)
Adult male Japanese vowels
Fu and Shannon(1999)
Identification accuracy follows an inverted U-shaped function.
Statistical modeling of perception
Extend the sliding template analogy to incorporate the prior probability.
Argumentation with equation (11)
Implicit prior distribution of Ψs
16.4 Estimation of Ψsinformed by g0
16.4.1 Method 4: One-shot plug-in substitution for Ψs
Equation(14)
Empirical correlation between speakers’ g0 ranges
Their format ranges
Figure 16.3
Information for the three datasets pooled
Equation(15)
Modify equation (7)
Closely related that of Miller(1989)
16.4.2 Method 5: MAP modulated by conditional probability of Ψs given g0
Ignores the error inherent in estimation of Ψs from g0
Equation(16)
16.4.3 Method 6: Joint maximization of P(v|Gt,Ψs ) and P (Ψs, g0t)
Represent the conditional probability ofΨs, given g0t via regression
Equation (21)
Is a complete expression for an estimate of the posterior mode ofΨs
In method 5 and 6
Provide the very convenient ways to
Modulate the effects of g0 on vowel identification parametrically

[音韻] Part three: Coarticulary Nasalization and the phonological development: Data from Italian and English Nasal=Fricative sequences Ch 11.3~4

Part three: Coarticulary Nasalization and the phonological development: Data from Italian and English Nasal-Fricative sequences Ch 11.3~4
11.3Method
Previous findings between oral air emission for the production of oral sound and the extend of the closure of the VP opening.
Positive correlation(Lubker and Moll 1995)
Current method
11.3.1 Speech material
Table 11.1 Words used in the experiment
The words are placed as in below
Italian: Dico X di
English: I said X again
And read five times by each subject
11.3.2 Procedure
Oral and nasal flows were transduced by
two-chamber Glottal Enterprise Rothenberg mask.
Audio signal were recorded by
A high-quality microphone attached to the exterior of the mask
11.3.3 Analysis
First analyzed with PCquitier.
Display of acoustic waveform
Spectrogram
Oral and nasal flow
When vowel is oral & nasal is fully articulated
When vowel is nasalized before a fully articulated nasal consonant
When vowel is nasalized before a weakly articulated nasal consonant
A fully nasalized vowels co-occurrence of nasal flow
11.3.4 Measures
11.3.4.1 Acoustic analysis
Duration measures were taken of
the test & control of Vs
nasalized portions of pre-nasal Vs, Ns,Fs
11.3.4.2 Nasal airflow
Difference at the nasal onset and offset
Figure 11.2
The interpretation of the nasal flow with the thresholds were label as
tN1, tN2, tN3, tN4,
The peak time of nasal flow was labeled t Npeak, -tN,
11.3.4.3 Oral airflow
Oral movement with a piecewise linear envelop
The envelop was used to compute the time lag
From the maximum of oral closure to the nasal peak
11.3.4.5 Statistical analysis
One-way ANOVAs
Acoustic data
Two-ways ANOVAs
Within-group
Between-group
By averaging, by group
Same to aerodynamic data
Averaging values across VNF, and VNTS context
11.4 Result
11.4.1. Acoustic analysis
Figure 1.1(left panel)
Typical case in N1 data
The vowel is heavily nasalized
The nasal consonant is weakly articulated before the following voiceless fricatives
Figure 1.1(right panel)
The release of
the oral occlusion for the nasal consonant between the velic closure(nasal peak)
The result of the acoustic analysis
Two-ways ANOVAs
Table 11.2-4
As expected
There is defect of vowel quality on vowel duration
And the duration of vowel nasalization
Table 5
N1 has the longest vowels and shortest oral consonants(F,TS) in
VNF
VNTS sequence

2008年6月5日 星期四

[音韻] Further reading of Ch3: introduction of Keith Johnson

http://linguistics.berkeley.edu/~kjohnson/Keith_Johnson_vita.html

http://www.amazon.com/Talker-Variability-Speech-Processing-Johnson/dp/0123865603

[音韻] Further reading of Ch6《Phonology & phonetic Evidence》 by Bruce Connel, University of Oxford

http://www.cambridge.org/catalogue/catalogue.asp?isbn=9780521483889

[音韻] Futher source of Ch11:《The phonetics & phonology of nasal gestures》University of Michigan

http://www.haskins.yale.edu/PowerPoint/Beddor_2007-09-27.ppt

[音韻] part two: Phonological Universals and the control and regulation of speech production ch 6.5~6.6

Part two: Phonological Universals and the control and regulation of speech production ch 6.5~6.6
The relation between Ps & F0
Experimental data shows
Ps & F0 are not well correlated
Changes of the glottis and of the vocal tract
Control of Ps & F0
By Laryngeal musculature
High level of Ps, f0 is elevated
Requirement
Produce trills depend on the aerodynamic conditions
Necessity
Building physiological models for consonants and for intonation
Studies like, Lieberman
Ladeforged
Collier
Maeda
Strik and Boves

[音韻] Part 1.Decisions and Mechanisms in Exemplar-based phonology ch 3.5

Part 1.Decisions and Mechanisms in Exemplar-based phonology ch 3.5
Recommendation
The cognitive basis of phonological knowledge to better understand the ecology of language.
Also shown on Ohala’s work on the phonetic and historical basis of language sound patterns.
Main goals of the chapter
Base on the traditional research in cognitive phonology
General approach to phonological modeling
Outlines of the answers
Must be in exemplars-based
In linguistics memory are from smaller phonetic/phonological units
Representation must be rich with phonetic details