Speech Analysis and Synthesis by Linear Prediction of the Speech Wave

By L. H. Anauer. 2000

How researchers see this paper

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

In Speech & Audio

Citations in the corpus

High Fidelity Neural Audio Compression (2022, neutral): “Low bitrate parametric speech and audio codecs have long been studied (Atal & Hanauer, 1971; Juang & Gray, 1982), but their quality has been severely limited.”
HiFi-Codec: Group-residual Vector quantization for High Fidelity Audio Codec (2023, background): “Low-bitrate parametric audio codecs have long been studied [14,15], but these models’ quality is very limited.”
Speech Resynthesis from Discrete Disentangled Self-Supervised Representations (2021, background): “Low bitrate parametric speech codecs have long been studied [27], but their quality has been severely limited.”
WORLD: A Vocoder-Based High-Quality Speech Synthesis System for Real-Time Applications (2016, background): “2.2 CheapTrick: Spectral Envelope Estimation Algorithm [16], [18] The spectral envelope is also an important parameter in speech processing, and Cepstrum and linear predictive coding (LPC) [23] are typical algorithms for determining it.”
LPCNET: Improving Neural Speech Synthesis through Linear Prediction (2018, background): “Low complexity parametric synthesis models such as low bitrate vocoders have existed for a long time [8, 9], but their quality has always been severely limited.”
Improving Opus Low Bit Rate Quality with Neural Speech Synthesis (2019, background): “Linear prediction is long known [25] to represent the spectral envelope of speech very well, and this enables the non-linear components of the network to focus on a spectrally flat excitation waveform.”
A Real-Time Wideband Neural Vocoder at 1.6 kb/s Using LPCNet (2019, background): “Very low bitrate parametric codecs have existed for a long time [1, 2], but their quality has always been severely limited.”
Neural Analysis and Synthesis: Reconstructing Speech from Self-Supervised Representations (2021, methodology): “We measured naturalness with 5-scale mean opinion score (MOS [1-5]).”
LVCNet: Efficient Condition-Dependent Modeling Network for Waveform Generation (2021, unknown): “In the traditional linear prediciton vocoder [16], a simple allpole linear filter is used to generate waveform in autoregressive way, of which the linear prediction coefficients is calculated according to the acoustic features.”
Introduction to Digital Speech Processing (2007, methodology): “In this case, the coefficients ki behave as reflection coefficients at the tube boundaries [3, 78, 101].”
Text-to-Speech Synthesis (2009, background): “Linear prediction is of cours e the basis of linear predictive coding (LPC) which was developed in the late 1960s and early 1970s [2 33], [20] [301] [294].”
Extraction of Visual Features for Lipreading (2002, methodology): “At that time, features for ASR were generally either derived from a model-based approach, linear prediction [2], or a data-driven approach using the short-term spectrum of the speech signal [31].”
CheapTrick, a spectral envelope estimator for high-quality speech synthesis (2015, methodology): “In particular, since spectral envelope estimation was important for speech synthesis, several estimation algorithms such as Cepstrum (Noll, 1964; Oppenheim, 1969) and linear predictive coding (LPC) (Atal and Hanauer, 1969) have been proposed.”
A spectral envelope estimation method based on F0-adaptive multi-frame integration analysis (2012, background): “Some other widely known spectral envelope representation methods use linear predictive coding (LPC) [5, 6], line spectral pairs (LSP), or cepstrum.”
Excitation source analysis for high-quality speech manipulation systems based on an interference-free representation of group delay with minimum phase response compensation (2014, methodology): “Source ﬁlter models provide a basis for VOCODER [12], LPC [13, 14], Cepstrum-based models [15, 16, 17] and STRAIGHT 1 [1, 2].”
PLATINUM: A method to extract excitation signals for voice synthesis system (2012, methodology): “Since sound quality of the synthesized voice depends on the estimation performance, many methods for accurately estimating each parameter have been proposed for high-quality voice synthesis [4–6].”
Discretization and Re-synthesis: an alternative method to solve the Cocktail Party Problem (2021, result): “Similar to the low-bitrate speech codecs [16]–[19], speech discretization aims at encode the speech input into a discrete sequence, shown in Fig.”