Isolating the energetic component of speech-on-speech masking with ideal time-frequency segregation.

By D. Brungart, Peter S. Chang, B. Simpson, Deliang Wang. Journal of the Acoustical Society of America, 2006

416 citations recorded by the source corpus.

Subfields: Cognitive Neuroscience, Speech and Hearing, Signal Processing

View Paper: https://www.semanticscholar.org/paper/76d4324408eca6b71afb6b3cb775ce7e1fa287cb

How researchers see this paper

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

In Cognitive Neuroscience

In Speech and Hearing

In Signal Processing

Citations in the corpus

Supervised Speech Separation Based on Deep Learning: An Overview (2017, neutral): “ideal binary masking dramatically improves speech intelligibility for normal-hearing (NH) and hearing-impaired (HI) listeners in noisy conditions [1], [16], [109], [173].”
Cocktail Party Processing via Structured Prediction (2012, methodology): “We adopt the common choice of LC = 0 in this paper [7].”
Binaural Classification for Reverberant Speech Segregation Using Deep Neural Networks (2014, background): “Recent speech perception research shows that IBM segregation produces large improvements of speech intelligibility in noise for normal-hearing listeners [6], [22], [3] and hearing-impaired listeners [2], [37].”
An Algorithm for Intelligibility Prediction of Time–Frequency Weighted Noisy Speech (2011, methodology): “This includes single-channel noise-reduction algorithms (see the work from Loizou [8] for an overview), but also speech separation techniques like ideal time frequency segregation (ITFS) [9], where typically a binary TF-weighting is used.”
A short-time objective intelligibility measure for time-frequency weighted noisy speech (2010, background): “, [8], but also speech separation techniques like ideal time frequency segregation (ITFS) [9], where typically a binary TF-weighting is”
A deep neural network for time-domain signal reconstruction (2015, neutral): “Note that for Eq.”
An Algorithm for Predicting the Intelligibility of Speech Masked by Modulated Noise Maskers (2016, methodology): “The fourth data set consists of the noisy signals from Additive Noise Set III , processed using the ITFS technique [36].”
Speech intelligibility in reverberation with ideal binary masking: effects of early reflections and signal-to-noise ratio threshold. (2013, methodology): “For a given input SNR, previous work has shown that the IBM improves intel-ligibility over a range of LC values (Brungart et al. , 2006, 2009; Kjems et al. , 2009).”
Exploring Monaural Features for Classification-Based Speech Segregation (2013, background): “A series of recent experiments [5], [24], [37] shows that IBM processing of sound mixtures yields large speech intelligibility gains.”
Pitch-based monaural segregation of reverberant speech. (2006, background): “Speech reconstructed from the ideal binary mask has been shown to be highly intelligible even when extracted from multisource mixtures and also to produce large improvements in robust speech recognition and human speech intelligibility Cooke et al., 2001; Roman et al., 2003; Brungart et al., 2006 .”
A Supervised Learning Approach to Monaural Segregation of Reverberant Speech (2009, background): “However, the harmonic structure is distorted by reverberation as reﬂections of each harmonic combine with the direct sound.”
A Tandem Algorithm for Pitch Estimation and Voiced Speech Segregation (2010, background): “Recent psychoacoustic studies provide strong evidence that the IBM leads to large improvements of human speech intelligibility in noise [9] [25].”
On Training Targets for Supervised Speech Separation (2014, background): “Furthermore, IBM processing has been shown to yield large speech intelligibility improvements even in extremely low SNR conditions [1], [2], [19], [33].”
Towards Scaling Up Classification-Based Speech Separation (2013, background): “Even if T-F unit features used for classification are robust to changing background noises, generalization to new speakers and SNRs is still an issue.”
Reasons why current speech-enhancement algorithms do not improve speech intelligibility and suggested solutions (2011, background): “Hence, it is not surprising that speech synthesized based on the IdBM criterion improves intelligibility [28], [29], [37].”
An algorithm to improve speech recognition in noise for hearing-impaired listeners. (2013, background): “Brungart et al. (2006) found that NH listeners could achieve near-perfect intelligibility in one- to three-talker interference.”