Creating models of
prototype matching in the context of human sensitivity of despoiled speech.
Reviewing models of auditory periphery by including the role
of the plunging pathway in making the cochlear response to speech sounds robust
to degradation in acoustic conditions.
Project’s attainment will contribute to following and have worth:
To develop a machine is
area of this research which will use state-of-the-art non-linear peripheral
auditory models (PAM) connected to a perceptually inspired prototype matching
to (1) foresee phonetic confusions made by normally listeners, and (2) expect fluency
of distorted speech generated by naturally spoken speech through realistic
C. Application of Cortical Processing Theory to Acoustical
The investigation began into the issue of short-time calculation
of the speech-based STI and the comparison of which with long-term computation
results. And, for the simple case of where we added noise i.e. noise addition, its
comparison with short-time SNR. To
understand, 0 dB SNR and 0 dB SNR plus sound audio conditions have been
considered. The speech-based STI
techniques explained above which are analyzed, include an envelope recession
method, a normalized correlation method and a normalized covariance method.
When speech plus noise is examined, all three above methods qualitatively track
short-time fluctuations in SNR as seen below in Figs 1 and 2
B.4. Short and Long Time Calculation of the Speech-Based STI
Selected initial data for frequency
modulation discrimination tested at 110, 139, 175, and 220 Hz is given in
Figure 2. The modulation envelope at the
tested frequency was used to modulate an 880 Hz sinusoidal transferor.
The prompt consisted of two tones, a reference
and a target. The reference frequency
was randomized within a 4 semitone interval.
So for the 440 Hz tone, the reference could be between 392 and 494 Hz. While
always the one with higher tone would be target. Both the target and reference were 500ms long
with a 200ms gap between. Note that
normal hearing performance on this task for non-musicians is approximately a
quarter of a semitone. The flyer
cochlear implant subject of Figure 1 at higher frequency showed comparatively
For Psychoacoustic procedures The
testing model used in initial data collection was two alternative forced choice models with a
two down, one up decision rule. Such a model converges to a 70.7% correct
response criteria. Selected initial data
for pure tone frequency discrimination tested at 440, 880, 1760 and 3520 Hz is
given in Figure 1.
Initial psychoacoustic data has been collected
for one subject having cochlear implant.
The emphasis of data was on pure tone frequency discrimination,
amplitude discrimination, and modulation frequency discrimination. In the first
MSI experiments involving subjects having cochlear implant, initial psychoacoustic
data will facilitate on making decision towards picking psychoacoustic and
speech reception testing battery. On software control and analysis functions for
collection of this data very generous work was done.
B2. Role of Discernment on
Performance by speech reception
For predicting speech reception by cochlear
implant users very much effort was done to develop STI and NCM models. Precisely,
for comparing speech reception scores with different cochlear implant users as
well as normal hearing ones part of model was developed listening. The
developed component is based on an efficiency factor that scales the standard values.
B.1 Further Development of the
STI and NCM Models.
B. Models of speech fluency
Although supra-threshold(a incentive that can produce action in cells) effects of hearing
impairment are widely believed to be related to the decreased firmness on psychoacoustic
tasks and on impaired listeners producing bad speech-reception abilities, the role of
reduced sound to be heard capacity itself in explaining the consequences of
hearing loss is as not fully implicit.
Work has begun on broad review of the literature on the effects of
hearing loss on speech-reception tasks to examine the evidence for
supra-threshold deficits. The review is organized into five major categories of
studies that include (a) temporal resolution, (b) intensity resolution, (c)
spectral resolution, (d) speech reception, and (e) correlational studies of
speech and psychoacoustic abilities.
A-4. Review of Earlier Research on the Role of Acoustics
in Expecting Effects of Hearing Impairment:
Algorithms for signal processing have been
developed for simulation of two types of hearing-aid processing. These processing modes include (a) linear
amplification for frequency-gain characteristic and (b) multi-band amplitude
compression amplitude based on a wide-dynamic range
A-3. Signal Processing for Simulation of Hearing
A Software has been developed for obtaining quantities of Speech Reception Threshold (SRT)
for HINT sentences as a function of the type of background noise and noise level. This procedure measures the
Speech-to babble (S/B) ratio where it is required for 50%-correct reception of
sentences using corresponding lists of phonemically stable materials for speech.
A-2. Speech Testing in Background Noise
The method which is currently
being implemented is (a), the use of additive-noise masking is limited to the
first 40 dB of hearing loss at a given frequency and the remaining loss is
simulated using (b) multi-band expansion. Using this way by combining both, we’ll
be able to simulate hearing loss in the range of mild to profound loss using
noise and the signal presentation levels that do not easy listening levels in
Signal-processing algorithms have
been implanted to produce results of hearing loss in those who have normal
hearing. MATLAB software is used in which these have been developed. And the
system is based on the amalgamation of
two varying ways towards hearing-loss simulation: (a) the addition of noise to raise inceptions of a normal-hearing
listeners so that it can be matched with
those of given hearing-impaired listener; and (b) the use of multi-band
expansion in which level-dependent mitigations are applied to sounds in
different frequency bands to map tone levels at the impaired listener’s
threshold to that of a normal listener
and to mimic the rapid growth of loudness observed in hearing loss.
A-1. Simulations of Sensorineural Hearing Loss
As hearing impaired people, have
issues towards speech reception so, the main aim of this research is related to
analyze the points responsible for this. And in result developing such methods
and techniques which can overcome these losses. Taking it to end if this is
successful, then it’ll be helpful in designing goals for improving the wearable
hearing aids and establishing new techniques for aural analysis and to
understand both acoustic function and speech perception problems.
Role of reduced acoustics:
To increase the effectiveness of
processing signal techniques that guarantee success for doing so.
To consider the relation of functional
characteristics of hearing impairments to reduced speech-reception capacity.
To assess effects of speech style
and inconsistency on its reception by impaired listeners.
To evaluate the effects of style of speech
articulation and variability in speech production on speech reception by
hearing impaired listeners.
Such model developments which can
easily predict the alterations of speech signal on articulacy
Focus point of this project goes
The goal of this research is to
develop a hearing aid for people suffering from sensorineural hearing
impairment and cochlear implants for impaired ones. Impaired listeners suffer
various transformations of speech signals on speech reception, which occurs due
to inadequate knowledge. The goal is specifically for to improve speech
reception, which comes from speed processing.