Diphone synthesis using an overlap-add technique for speech waveforms concatenation

By Francis Charpentier, M. Stella. 1986

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

Neural Speech Synthesis with Transformer Network (2018, methodology): “and stitch them by some algorithms such as Viterbi (Viterbi 1967) followed by signal process methods (Charpentier and Stella 1986; Verhelst and Roelands 1993) to generate new waves.”
TTS-Portuguese Corpus: a corpus for speech synthesis in Brazilian Portuguese (2020, methodology): “Several acoustic models have been proposed such as the classical formant model [7], Linear Prediction Coefficients (LPC) model, the Pitch Synchronous Overlap and Add (PSOLA) models [8] widely used in TTS engines like Microsoft Speech API.”
Compression Artifacts in Perceptual Audio Coding (2008, background): “Table I lists these artifacts in terms of the perception, the generation source, the time/frequency feature, the music types, the typical tracks and the existing relief methods.”
Controllable Neural Prosody Synthesis (2020, methodology): “In Table 3, we see that our PS-WaveNet significantly outperforms the naturalness of WORLD while achieving comparable performance to PSOLA.”
End-To-End Speech Synthesis Applied to Brazilian Portuguese (2020, methodology): “Several models have been used to implement this block, such as the classical formants model [Klatt, 1980], Linear Prediction Coefﬁcients (LPC) model, the Pitch Synchronous Overlap and Add (PSOLA) models [Charpentier and Stella, 1986] largely used in several TTS engines like Microsoft Speech API.”
Introduction to Digital Speech Processing (2007, methodology): “Various smoothing methods based on the concepts of time domain harmonic scaling (TDHS) [76] and pitch synchronous overlap add (PSOLA) [20, 31, 82] have been proposed and optimized for such smoothing/modification at the boundaries between adjacent diphone units.”
A survey on speech synthesis techniques in Indian languages (2020, background): “In this type of synthesizers, all the diphones (sound-to-sound transitions) occurring in a language are contained in the speech database [112] requiring a memory of around 3–4 Megabytes.”