Assessment of Vocal Noise via Bi-directional Long-term Linear - - PowerPoint PPT Presentation

Assessment of Vocal Noise via Bi-directional Long-term Linear Prediction of Running Speech

• F. Bettens*, F. Grenez*, J. Schoentgen*,**

*Université Libre de Bruxelles **National Fund for Scientific Research

Belgium

Vocal Dysperiodicities Cause Pitch Breaks, Phonation Breaks, Timbre Breaks, … Transients Vibrations Ventricular Folds or Ary-Epiglottic Ligaments, … “Parasitic” Vibrations Breathiness, Breathy Voice, Whispery Voice, … (Audible) Additive Noise Owing to Turbulence Vocal Jitter & Shimmer, Frequency & Amplitude Tremor Perturbations Diplophonia, Bi-Phonation, Random Vibrations Vocal Fold Dynamics

Existing Cues of Vocal Noise

• Detection of individual vocal cycles

(or harmonics)

Steady vowel fragments (Pseudo)-Periodicity

Period Perturbation Quotient Amplitude Perturbation Quotient Harmonics-to-Noise Ratio

Objectives : Analyses of Dysperiodicities

• Give up request that speech fragments are :

(Pseudo)-Periodic Steady

• Any Speech Fragment :

Modal Voices & (Very) Hoarse Voices Sustained Vowels & Running Speech

Motivation : Analysis of Running Speech

• Voicing in running speech

Variable acoustic impedance Voicing onsets & offsets Variable pressure drops Variable laryngeal positions

• Voice Loading

Double Linear Predictive Analysis

• Conventional short-term linear prediction:
• Long-term linear prediction:

remove existing correlations ⇒ unpredictable noise component (Qi, 1999)

] [ ] [ ] [ ] [ ] [

1

n x' n x n e i n x a n x'

S N i i

− = ⇒ − − = ∑

=

forward short-term prediction error

] [ ] [ ] [ ] [ ] [ n y' n y n e i P n y b n y'

L M i i

− = ⇒ − − − = ∑

=

forward double prediction error

] [ ] [ with n e n y

S

= signal speech : ] [ with n x

Double Linear Predictive Analysis Drawbacks:

– eS[n] is an artificial signal – the dysperiodicities in weighted sum x′[n] are omitted – eL[n] is inflated to the right of unvoiced/voiced boundaries ] [ ] [ ] [ ] [ ] [

1

n x' n x n e i n x a n x'

S N i i

− = ⇒ − − = ∑

=

] [ ] [ ] [ ] [ ] [ n y' n y n e i P n y b n y'

L M i i

− = ⇒ − − − = ∑

=

⇒ Solutions:

⇒ remove short-term linear predictive analysis stage ⇒ proceed to bi-directional analysis

• Forward long-term linear prediction:
• Backward long-term linear prediction:
• Bi-directional long-term linear prediction:

keep the “best”

(frame by frame)

Bi-directional Long-term Prediction

( )

] [ , ] [ min ] [ n e n e n e

L L L

• =

⇒

forward long-term prediction error

] [ ] [ ] [ ] [ ] [ n ' y n y n e i P n y b n ' y

L M i i

• −

= ⇒ + + − = ∑

=

backward long-term prediction error

signal speech : ] [ ] [ with n x n y = ] [ ] [ ] [ ] [ ] [ n ' y n y n e i P n y b n ' y

L M i i

• −

= ⇒ − − − = ∑

=

bi-directional long- term prediction error

Long-term Prediction Distance : P Maximum of the auto-correlation function

example: steady vowel [a] (dysphonic speaker)

⇒ P = 184 (2 cycles)

↓

Vocal Noise Cue

Signal-to-Dysperiodicity Ratio:

SDR = 31,2 dB

speech signal dysphonic speaker bi-directional long-term prediction error

SDR = 10,1 dB

healthy speaker x[n] eL[n]

example: steady vowel [a]             − =

∑ ∑

= =

1 ] [ ] [ log 10

1 2 1 2 10

P P

N n L N n dB

n e n x SDR

Results1:Sentence(1 female speaker; modal phonation type) (http://www.limsi.fr/VOQUAL/ : “Il est sorti avant le jour”)

speech signal forward long-term prediction error bi-directional long-term prediction error

segments [il]

Results 2 : Sentence (1 female speaker; 5 phonation types) (http://www.limsi.fr/VOQUAL/ : “Il est sorti avant le jour”)

Double prediction Long-term prediction Direction Signal SDR (dB) SDR (dB) Modal 25.7 19.5 Rough1 16.9 11.4 Rough2 13.9 8.0 Rough3 9.8 3.6 Bi-directional Whisper 9.5 3.2 Modal 25.4 16.2 Rough1 16.8 10.3 Rough2 13.7 6.9 Rough3 9.6 2.7 Forward Whisper 9.3 1.8

SDR bi-directional > SDR forward SDR double > SDR long-term

Conclusion

The forward & backward long-term prediction

• f speech enables the analysis of any speech

signal with a view to the assessment of the vocal noise (i.e. vocal dysperiodicities) The analysis is not based on any assumptions regarding the periodicity or stationarity of the speech signals