A short-time objective intelligibility measure for time-frequency weighted noisy speech

By Cees H. Taal, Richard Christian Hendriks, Richard Heusdens, Jesper Højvang Jensen. 2010

1,274 citations recorded by the source corpus.

Subfields: Biomedical Engineering, Cognitive Neuroscience, Signal Processing

How researchers see this paper

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

In Biomedical Engineering

In Cognitive Neuroscience

In Signal Processing

Citations in the corpus

End-to-End Neural Speech Coding for Real-Time Communications (2022, neutral): “In evaluation, except the subjective listening test in Section 4.2, three metrics are used for ablation studies in Sec-Fig [20], STOI [21] and the DNSMOS [22].”
InstructTTS: Modelling Expressive TTS in Discrete Latent Space With Natural Language Style Prompt (2023, methodology): “For speech quality, we adopt Mel-cepstral distorion (MCD) [66], structural similarity index measure (SSIM) [67] and Short-Time Objective Intelligibility (STOI) [68] to evaluate the speech quality.”
Looking to listen at the cocktail party (2018, neutral): “2001], STOI [Taal et al. 2010] and SDR from the BSS eval toolbox [Vincent et al.”
Composition of Deep and Spiking Neural Networks for Very Low Bit Rate Speech Coding (2016, methodology): “ity. This approach was found to provide good predictions of subjective intelligibility scores for speech coding as used in cellular networks [71]. We used the publicly available MATLAB implementation [72]. TABLE V The STOI intelligibility evaluation ofHMM coding operating at 213 bit/s, NN coding ope”
SDR – Half-baked or Well Done? (2018, methodology): “While the merits of various objective measures such as PESQ [12], Loizou’s composite measure [13], PEMO-Q [14], PEASS [15], or STOI [16], could be debated and compared, we are concerned here with an issue with the way the widely relied upon BSS eval tool-box [17] has been used.”
ProDiff: Progressive Fast Diffusion Model for High-Quality Text-to-Speech (2022, methodology): “For objective evaluation, ProDiff also demonstrates the outperformed performance in MCD, PESQ, and STOI, superior to all baseline models.”
Bilateral Denoising Diffusion Models (2021, positive): “In addition, to measure the noisiness and the distortion of the generated speech relative to the reference speech, we adopted two commonly used metrics in speech enhancement – the perceptual evaluation of speech quality (PESQ) [44] and the short-time objective intelligibility (STOI) measure [45].”
DeepMMSE: A Deep Learning Approach to MMSE-Based Noise Power Spectral Density Estimation (2020, methodology): “The short-time objective intelligibility (STOI) measure [55], [56] is used to evaluate the objective speech intelligibility.”
Learning Individual Speaking Styles for Accurate Lip to Speech Synthesis (2020, methodology): “We compare our models with the previous lip to speech works using three standard speech quality metrics: Short-Time Objective Intelligibility (STOI) [32] and Extended ShortTime Objective Intelligibility (ESTOI) [17] for estimating the intelligibility and Perceptual Evaluation of Speech Quality (PESQ”
Infergrad: Improving Diffusion Models for Vocoder by Considering Inference in Training (2022, positive): “For objective measurements, we employ log-mel spectrogram mean squared error (LS-MSE), multi-resolution STFT (MRSTFT) [26], perceptual evaluation of speech quality (PESQ) [30] and short-time objective intelligibility (STOI) [31].”
Fréchet Audio Distance: A Metric for Evaluating Music Enhancement Algorithms (2018, background): “…which are not trivially applicable to evaluating music enhancement approaches, are automatic metrics such as Perceptual evaluation of speech quality (PESQ) [16] and short-time objective intelligibility (STOI) [19] that approximate perceptual level metrics without requiring any human raters.”
On Mean Absolute Error for Deep Neural Network Based Vector-to-Vector Regression (2020, methodology): “A higher STOI score corresponds to a better speech intelligibility.”
Deep Residual-Dense Lattice Network for Speech Enhancement (2020, methodology): “STOI is the short-time objective intelligibility measure (in %) (Taal et al. 2010).”
Speech Enhancement of Noisy and Reverberant Speech for Text-to-Speech (2018, background): “In contrast to what PESQ and STOI predicted (see Table III), the samples from RNN-N (V) were rated superior to noisy speech.”
Psychoacoustic Calibration of Loss Functions for Efficient End-to-End Neural Audio Coding (2020, background): “Other neural speech enhancement systems implement short-time objective intelligibility (STOI) [16] and perceptual evaluation of speech quality (PESQ) [17] as the loss [18], [19].”
Avocodo: Generative Adversarial Network for Artifact-free Vocoder (2022, methodology): “To measure the perceived quality of the synthesized speech, the mel-cepstral distortion (MCD)[32], structural similarity index (SSIM)[33], perceptual evaluation of speech quality (PESQ)[34] and short-time objective intelligibility (STOI)[35] were calculated.”
Time-Frequency Masking-Based Speech Enhancement Using Generative Adversarial Network (2018, methodology): “Moreover, to show the improvement in speech intelligibility, we also calculated Short-Time Objective Intelligibility (STOI) measure [27].”
A Fully Convolutional Neural Network for Speech Enhancement (2016, methodology): “In addition, Short time Objective Intelligibility (STOI) [26] and Perceptual Evaluation of Speech Distortion (PESQ) [27] — both assume that human perception has short term memory and hence the error is measured nonlinearly in time of interest— were used to measure the subjective quality of…”