Objective measures of speech quality

By S. Quackenbush. 1995

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Speech Denoising with Deep Feature Losses (2018, background): “As baselines, we use a Wiener filtering pipeline with a priori noise SNR estimation (as implemented in [39]), and two recent state-of-the-art methods that use deep networks to perform end-to-end denoising directly on the raw waveform: the Speech Enhancement Generative Adversarial Network (SEGAN) [21”
SEGAN: Speech Enhancement Generative Adversarial Network (2017, neutral): “• SSNR: Segmental SNR [35, p. 41] (from 0 to ∞ ).”
Evaluation of objective measures for speech enhancement (2006, methodology): “Composite objective measures are obtained by linearly combining existing objective measures to form a new measure [1].”
Noise estimation by minima controlled recursive averaging for robust speech enhancement (2002, background): “Third, the strong correlation of speech presence in consecutive frames is utilized (via ).”
A noise-estimation algorithm for highly non-stationary environments (2006, methodology): “Two additional objective measures were used to evaluate and compare the performance of the proposed noise-estimation algorithm: the segmental SNR and the log-likelihood ratio (LLR) measure (Quackenbush et al., 1988).”
Speech intelligibility in reverberation with ideal binary masking: effects of early reflections and signal-to-noise ratio threshold. (2013, background): “…other recent work (Loizou and Kim, 2011), have helped to illustrate why measures of algo-rithm performance based on signal-to-noise ratio (SNR) (Quackenbush et al. , 1988; Greenberg et al. , 1993; ANSI, 1997) do not correlate well with the intelligibility of processed speech (Hu and Loizou,…”
Speech Quality Assessment (2011, methodology): “Several methods for evaluating speech quality have been proposed in the literature [1].”
Separate Anything You Describe (2023, unknown): “…apply the perceptual evaluation of speech quality (PESQ) [72], mean opinion score (MOS) predictor of signal distortion (CSIG), MOS predictor of background-noise intrusiveness (CBAK), MOS predictor of overall signal quality (COVL) [73] and segmental signal-to-noise ratio (SSNR) [74] for evaluation.”
Speech Enhancement for Noise-Robust Speech Synthesis Using Wasserstein GAN (2019, methodology): “The gain in SNR through the enhancement processing is being evaluated by segmental SNR (SSNR) [23].”
Techniques for Harmonic Sinusoidal Coding (1997, background): “The SD for a given frame is defined as [59]:”
Noise power spectral density estimation based on optimal smoothing and minimum statistics (2001, methodology): “The standard error (s.e.) of the test was about 0.9 DAM points.”
Reasons why current speech-enhancement algorithms do not improve speech intelligibility and suggested solutions (2011, methodology): “The time-domain segmental (and overall) SNR measure has been used widely and frequently for evaluating speech quality in speech coding and enhancement applications [10], [11].”
Evaluation of Objective Quality Measures for Speech Enhancement (2008, methodology): “Among all objective measures considered, the PESQ measure is the most complex to compute and is the one recommended by ITU-T for speech quality assessment of 3.2 kHz (narrow-band) handset telephony and narrow-band speech codecs [7], [8].”
Qualinet White Paper on Definitions of Quality of Experience (2013, background): “Once being identified, the features can also be scaled directly in a psychophysical experiment, see e.g. Voiers (1977), Quackenbush et al. (1988) or Wältermann et al. (2012).”
A two-stage algorithm for one-microphone reverberant speech enhancement (2006, background): “Many objective speech quality measures (for example, see [ 30 ]) are solely based on the magnitude spectrum of speech, in part motivated by the study of Wang and Lim [33] showing that phase distortion is not important for enhancing speech mixed with a moderate level of white noise.”
Improved Speech Enhancement with the Wave-U-Net (2018, unknown): “Last is the Segmental Signal-to-Noise Ratio (SSNR) [10] (from 0 to ∞ ).”
Deep Learning for Environmentally Robust Speech Recognition (2017, unknown): “For speech processing, 1D ﬁlters instead of 2D ﬁlters are particularly considered when directly addressing the raw signals.”