Discrete-Time Speech Signal Processing: Principles and Practice

By Thomas F. Quatieri. 2001

1,090 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/9ff267fd80642576059821dcf7486a237b90516e

How researchers see this paper

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

1. Xu Tan
2. Li Deng
3. Junichi Yamagishi
4. Jinyu Li
5. Dong Yu
6. Eunwoo Song
7. Ryuichi Yamamoto
8. Frank K. Soong

Citations in the corpus

• A Reverberation-Time-Aware Approach to Speech Dereverberation Based on Deep Neural Networks (2017, methodology): “The autocorrelation of y(n) is given as follows [31],”
• Voice Morphing that improves TTS quality using an optimal dynamic frequency warping-and-weighting transform (2016, methodology): “where ωh is the radial frequency of the h-th sinusoid, the nominator corresponds to the power of the morphed spectra and the denominator corresponds to the power of the unmorphed spectra, according to the Parseval theorem [30].”
• Spoofing and countermeasures for speaker verification: A survey (2015, background): “Since the human auditory system is thought to be relatively insensitive to phase (Quatieri, 2002), vocoders typically do not reconstruct speech-like phase information.”
• TTS-by-TTS: TTS-Driven Data Augmentation for Fast and High-Quality Speech Synthesis (2020, methodology): “In this vocoder, the speech waveform is generated by the WaveNet-based mixture density network [12] within the framework of the human speech production mechanism [13].”
• Vocaine the vocoder and applications in speech synthesis (2015, background): “The commercial interest for vocoders started with speech coding, e.g. the Sinusoidal Transform Codec (STC) [2, 3] and Advanced MultiBand Excitation (AMBE) [4] but was later reduced for about a decade following the dominance of CELP codecs.”
• Neural Homomorphic Vocoder (2020, background): “The group delay of filters affects the timbre of speech, as reported in several papers [21, 24, 25] and books [26, 27].”
• The importance of phase in speech enhancement (2011, methodology): “The enhanced speech signal, y(n), is constructed by taking the inverse STFT of Y(n, k) followed by least-squares overlap-add synthesis (Griffin and Lim, 1984; Quatieri, 2002).”
• A review of depression and suicide risk assessment using speech analysis (2015, methodology): “The foundation for feature extraction methodologies are not covered in this section; for this the reader is referred to either Kinnunen and Li (2009) or Quatieri (2001).”
• Effective Spectral and Excitation Modeling Techniques for LSTM-RNN-Based Speech Synthesis Systems (2017, background): “Accordingly, many researchers have analyzed the effects of various vocoding techniques on the quality of synthesized speech [5]–[8], as the speech analysis/synthesis module itself determines the best achievable quality of text-to-speech (TTS) systems.”
• ExcitNet Vocoder: A Neural Excitation Model for Parametric Speech Synthesis Systems (2018, background): “In the typical source-filter theory of speech production [5], the residual signal, i.”
• An Overview of Noise-Robust Automatic Speech Recognition (2014, background): “which is referred to as the Wiener filter [106], [107], and can be realized only if Sxx(f) and Snn(f) are known.”
• Spoken Language Recognition: From Fundamentals to Practice (2013, background): “The early efforts of acoustic–phonetic approaches to language recognition probably began in the 1980s.”
• Unsupervised Speech Decomposition via Triple Information Bottleneck (2020, background): “The Source-Filter Model Early research on speech generation proposed the source-flter model (Quatieri, 2006), and many subsequent research efforts try to decompose speech into the source that includes pitch and the flter that includes content, using signal processing approaches, such as linear…”