Speaking rate normalization with lattice-based context-dependent phoneme duration modeling for personalized speech recognizers on mobile devices

2013 ◽  
Author(s):  
Ching-Feng Yeh ◽  
Hung-yi Lee ◽  
Lin-shan Lee
Keyword(s):  
Mobile Devices ◽  
Speaking Rate ◽  
Speech Recognizers ◽  
Context Dependent ◽  
Duration Modeling

Context dependent modelling approaches for hybrid speech recognizers

2010 ◽  
Author(s):  
Alberto Abad ◽  
Thomas Pellegrini ◽  
Isabel Trancoso ◽  
João Neto
Keyword(s):  
Speech Recognizers ◽  
Context Dependent ◽  
Modelling Approaches

Rate-dependent Perception of Aspiration and Pre-aspiration in Icelandic

1996 ◽  
Vol 49 (3) ◽  
pp. 745-764 ◽  
Author(s):  
Jörgen Pind
Keyword(s):  
Voice Onset Time ◽  
Onset Time ◽  
Speaking Rate ◽  
Rate Dependent ◽  
The Face ◽  
Speech Cues ◽  
Context Dependent ◽  
Offset Time ◽  
Speech Segments ◽  
Over Time

Speech segments are highly context-dependent and acoustically variable. One factor that contributes heavily to the variability of speech is speaking rate. Some speech cues are temporal in nature—that is, the distinctions that they signify are defined over time. How can temporal speech cues keep their distinctiveness in the face of extrinsic transformations, such as those wrought by different speaking rates? This issue is explored with respect to the perception, in Icelandic, of Voice Onset Time as a cue for word-initial stop voicing, wordinitial aspiration as a cue for [h], and Voice Offset Time as a cue for pre-aspiration. All the speech cues show rate-dependent perception though to different degrees, with Voice Offset Time being most sensitive to rate changes and Voice Onset Time least sensitive. The differences in the behaviour of these speech cues are related to their different positions in the syllable.

Automated Acoustic Analysis of Oral Diadochokinesis to Assess Bulbar Motor Involvement in Amyotrophic Lateral Sclerosis

2020 ◽  
Vol 63 (1) ◽  
pp. 59-73 ◽  
Author(s):  
Panying Rong
Keyword(s):  
Amyotrophic Lateral Sclerosis ◽  
Acoustic Analysis ◽  
Speaking Rate ◽  
Disease Stage ◽  
Healthy Controls ◽  
Tongue Movement ◽  
Major Barrier ◽  
Syllable Repetition ◽  
Oral Diadochokinesis ◽  
Lateral Sclerosis

Purpose The purpose of this article was to validate a novel acoustic analysis of oral diadochokinesis (DDK) in assessing bulbar motor involvement in amyotrophic lateral sclerosis (ALS). Method An automated acoustic DDK analysis was developed, which filtered out the voice features and extracted the envelope of the acoustic waveform reflecting the temporal pattern of syllable repetitions during an oral DDK task (i.e., repetitions of /tɑ/ at the maximum rate on 1 breath). Cycle-to-cycle temporal variability (cTV) of envelope fluctuations and syllable repetition rate (sylRate) were derived from the envelope and validated against 2 kinematic measures, which are tongue movement jitter (movJitter) and alternating tongue movement rate (AMR) during the DDK task, in 16 individuals with bulbar ALS and 18 healthy controls. After the validation, cTV, sylRate, movJitter, and AMR, along with an established clinical speech measure, that is, speaking rate (SR), were compared in their ability to (a) differentiate individuals with ALS from healthy controls and (b) detect early-stage bulbar declines in ALS. Results cTV and sylRate were significantly correlated with movJitter and AMR, respectively, across individuals with ALS and healthy controls, confirming the validity of the acoustic DDK analysis in extracting the temporal DDK pattern. Among all the acoustic and kinematic DDK measures, cTV showed the highest diagnostic accuracy (i.e., 0.87) with 80% sensitivity and 94% specificity in differentiating individuals with ALS from healthy controls, which outperformed the SR measure. Moreover, cTV showed a large increase during the early disease stage, which preceded the decline of SR. Conclusions This study provided preliminary validation of a novel automated acoustic DDK analysis in extracting a useful measure, namely, cTV, for early detection of bulbar ALS. This analysis overcame a major barrier in the existing acoustic DDK analysis, which is continuous voicing between syllables that interferes with syllable structures. This approach has potential clinical applications as a novel bulbar assessment.

Sign in / Sign up

Export Citation Format

Share Document