Sara Akbarzadeh

Ms. Sara Akbarzadeh


Research Focus & Metrics

Academic Career Length

8Years (2019-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
27 May 2022 in IEEE Access
Verified
Journal Article
Open access
SASara Akbarzadeh
Sara Akbarzadeh
ELEdward Lobarinas
Edward Lobarinas
NKNasser Kehtarnavaz
Nasser Kehtarnavaz
A key function of modern hearing aids is compression or mapping of sound to the residual hearing range of those suffering from hearing loss. This paper presents a machine learning approach to personalize compression in hearing aids in an online manner. The online feature of this approach allows it to be deployed in the field. The significance of this personalized compression lies in enabling preferred hearing outcomes relative to the one-size-fits-all prescriptive compression rationales that are currently being used. This personalization approach utilizes maximum likelihood inverse reinforcement learning to establish a model of a hearing aid user’s preference based on paired comparisons by the user. The results of the preference paired comparisons between the personalized and standard prescriptive settings from ten subjects indicated that personalized settings were preferred about 10 times more than the standard prescriptive settings. In addition, a word recognition comparison was conducted showing that the personalized settings had no adverse impact on speech understanding in either quiet or in competing noise conditions.

12 August 2022 in Sensors
Verified
Journal Article
Open access
PDF file
ANAoxin Ni
Aoxin Ni
SASara Akbarzadeh
Sara Akbarzadeh
ELEdward Lobarinas
Edward Lobarinas
NKNasser Kehtarnavaz
Nasser Kehtarnavaz
Adaptive dynamic range optimization (ADRO) is a hearing aid fitting rationale which involves adjusting the gains in a number of frequency bands by using a series of rules. The rules reflect the comparison of the estimated percentile occurrences of the sound levels with the audibility and comfort hearing levels of a person suffering from hearing loss. In the study reported in this paper, a previously developed machine learning method was utilized to personalize the ADRO fitting in order to provide an improved hearing experience as compared to the standard ADRO hearing aid fitting. The personalization was carried out based on the user preference model within the framework of maximum likelihood inverse reinforcement learning. The testing of ten subjects with hearing loss was conducted, which indicated that the personalized ADRO was preferred over the standard ADRO on average by about 10 times. Furthermore, a word recognition experiment was conducted, which showed that the personalized ADRO had no adverse impact on speech understanding as compared to the standard ADRO.

1 September 2021 in Speech Communication
Verified
Journal Article
Closed access
SASara Akbarzadeh
Sara Akbarzadeh
SLSungmin Lee
Sungmin Lee
FCFei Chen
Fei Chen
CTChin-Tuan Tan
Chin-Tuan Tan

1 July 2019
Verified
Conference Paper
Closed access
SASara Akbarzadeh
Sara Akbarzadeh
SLSungmin Lee
Sungmin Lee
FCFei Chen
Fei Chen
CTChin-Tuan Tan
Chin-Tuan Tan
Most noise reduction algorithms in current assistive hearing devices were designed to reduce noise to achieve the best signal to noise ratio (SNR) possible, presuming that the perceived quality of speech can be compromised as long as the level of noise is relatively much lower than the level of speech. In this study, we investigated and compared various factors other than SNR that may possibly affect the quality rating pattern of normal hearing (NH), cochlear implant (CI) and hearing aid (HA) subjects. In one of the comparisons, we found that the perceived sound quality ratings of noisy sentences at the same SNR were dependent on the level of speech present in the sentences rather than the noise. Noisy sentences with a higher level of speech were rated higher in perceived sound quality by NH subjects. However, in a similar comparison, CI subjects rated higher in perceived sound quality when the level of speech present was lower, but HA subjects do not show any clear trend with their perceived sound quality ratings on the same sentences.

Verified
Journal Article
Open access
PDF file
SASara Akbarzadeh
Sara Akbarzadeh
SLSungmin Lee
Sungmin Lee
CTChin-Tuan Tan
Chin-Tuan Tan
In multi-speaker environments, cochlear implant (CI) users may attend to a target sound source in a different manner from normal hearing (NH) individuals during a conversation. This study attempted to investigate the effect of conversational sound levels on the mechanisms adopted by CI and NH listeners in selective auditory attention and how it affects their daily conversation. Nine CI users (five bilateral, three unilateral, and one bimodal) and eight NH listeners participated in this study. The behavioral speech recognition scores were collected using a matrix sentences test, and neural tracking to speech envelope was recorded using electroencephalography (EEG). Speech stimuli were presented at three different levels (75, 65, and 55 dB SPL) in the presence of two maskers from three spatially separated speakers. Different combinations of assisted/impaired hearing modes were evaluated for CI users, and the outcomes were analyzed in three categories: electric hearing only, acoustic hearing only, and electric + acoustic hearing. Our results showed that increasing the conversational sound level degraded the selective auditory attention in electrical hearing. On the other hand, increasing the sound level improved the selective auditory attention for the acoustic hearing group. In the NH listeners, however, increasing the sound level did not cause a significant change in the auditory attention. Our result implies that the effect of the sound level on selective auditory attention varies depending on the hearing modes, and the loudness control is necessary for the ease of attending to the conversation by CI users.