CITATION — REFERENCE ENTRY
Time-frequency masking for speech separation and its potential for hearing aid design — Trends in Amplification
- Key
- wang2008-tf-masking
- Authors
- Wang, DeLiang
- Issued
- 2008-12
- Type
- article-journal
- Container
- Trends in Amplification
- Volume
- 12
- Issue
- 4
- Pages
- 332-353
Raw CSL JSON
{
"DOI": "10.1177/1084713808326455",
"URL": "https://pmc.ncbi.nlm.nih.gov/articles/PMC4111459/",
"PMID": "18974204",
"page": "332-353",
"type": "article-journal",
"PMCID": "PMC4111459",
"issue": "4",
"title": "Time-frequency masking for speech separation and its potential for hearing aid design",
"author": [
{
"given": "DeLiang",
"family": "Wang"
}
],
"issued": {
"date-parts": [
[
2008,
12
]
]
},
"volume": "12",
"container-title": "Trends in Amplification"
}
Claims
-
IBM processing makes the intelligibility performance of hearing-impaired and normal-hearing listeners comparable.
"A striking conclusion from Wang et al.’s study is that IBM processing makes the intelligibility performances for HI listeners and NH listeners comparable."
-
Wang and colleagues suggested the ideal binary mask as a major computational goal of CASA, usable as an explicit criterion for evaluating CASA systems.
"Wang and colleagues suggested the notion of an ideal binary mask (IBM) as a major computational goal of CASA (Hu & Wang, 2001; Roman, Wang, & Brown, 2003; Wang, 2005), which can then be used as an explicit criterion for evaluating a CASA system."
-
IBM processing provides large intelligibility gains, larger for hearing-impaired than for normal-hearing listeners, and for modulated than for steady noise.
"First, IBM processing provides large intelligibility gains. Second, the intelligibility benefit is even larger for HI listeners than for NH subjects, and for modulated noise than for steady noise."
-
As of 2008, no study of time-frequency masking had targeted the hearing aid application directly.
"None of the studies on T-F masking have so far targeted the hearing aid application directly."
-
Time-frequency masking originated in computational auditory scene analysis and performs separation in the time-frequency domain.
"Originated in the field of computational auditory scene analysis, T-F masking performs separation in the time–frequency domain."
Available in