The effect of speech material on the band importance function for Mandarin Chinese

Abstract

Speech material influences the relative contributions of different frequency regions to intelligibility for English. In the current study, whether a similar effect of speech material is present for Mandarin Chinese was investigated. Speech recognition was measured using three speech materials in Mandarin, including disyllabic words, nonsense sentences, and meaningful sentences. These materials differed from one another in terms of the amount of contextual information and word frequency. The band importance function (BIF), as defined under the Speech Intelligibility Index (SII) framework, was used to quantify the contributions across frequency regions. The BIFs for the three speech materials were estimated from 16 adults who were native speakers of Mandarin. A Bayesian adaptive procedure was used to efficiently estimate the octave-frequency BIFs for the three materials for each listener. As the amount of contextual information increased, low-frequency bands (e.g., 250 and 500 Hz) became more important for speech recognition, consistent with English. The BIF was flatter for Mandarin than for comparable English speech materials. Introducing the language-and material-specific BIFs to the SII model led to improved predictions of Mandarin speech-recognition performance. Results suggested the necessity of developing material-specific BIFs for Mandarin. V

Simulator Verification Is Potentially Beneficial for the Fitting of Softband Bone Conduction Hearing Devices in Young Children

“…At each frequency, the sound energy was filtered through the 1/3 octave band filter centered at that frequency. The SII was calculated as the sum of audible sounds at each frequency multiplied by the frequency importance function (FIF), which weighted the contribution of speech at each frequency in the target frequency regions (21). The speech intelligibility of the BCHD output via the simulator was evaluated before and after the reprogramming.…”

Spectral weighting for sentence recognition in steady-state and amplitude-modulated noise

“…The qBIF procedure returns two aspects of the fitted SII model: (1) the BIF, with one spectral weight estimated for each frequency band in the filter bank and (2) a transfer function that links the SII value to proportion correct in keyword recognition. Following Du et al (2019) , the transfer function is defined as a logistic function, where θ and β are two parameters that correspond to the SII value associated with 50% keyword recognition and how rapidly the recognition performance increases with increasing SII values, respectively.…”

“…The BIF in the SII model is not invariant; it varies with features of the target speech, such as the type of speech material ( DePaolis, 1996 ; Du et al , 2019 ), the availability of visual information ( Bernstein and Grant, 2007 ; Bernstein et al , 2020 ), and presentation level ( Calandruccio et al , 2016 ; Shen et al , 2020 ). Recent research also suggests that the BIF may vary from individual to individual ( Shen and Kern, 2018 ; Varnet et al , 2019 ; Yoho and Bosen, 2019 ; Shen et al , 2020 ).…”

The impact of untreated hearing loss on depression, anxiety, stress, and loneliness in tonal language-speaking older adults in China

“…The perception of tone primarily depends on the fundamental frequency (F0) contour. On the other hand, in English, the F0 only impacts the intonation, which will not change the meaning of the word ( Du et al, 2019 ). Furthermore, high frequencies (6–8 kHz) show a decrease in sensitivity at least a decade earlier than for the speech frequencies ( Salvi et al, 2018 ), and hearing ability at these frequencies impacts speech perception ( Feng et al, 2010 ).…”