Speaker Adaptation in DNN-Based Speech Synthesis Using d-Vectors

By Rama Doddipatla, Norbert Braunschweiler, Ranniery da S. Maia. 2017

47 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Speech & Audio, 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 Artificial Intelligence

1. Xu Tan
2. Tao Qin
3. Tie-Yan Liu
4. Hung-yi Lee
5. Yu Zhang
6. Frank K. Soong
7. Ye Jia
8. Junichi Yamagishi

In Speech & Audio

1. Xu Tan
2. Hung-yi Lee
3. Tao Qin
4. Ye Jia
5. Tie-Yan Liu
6. Yu Zhang
7. Junichi Yamagishi
8. Yonghui Wu

In Signal Processing

1. Xu Tan
2. Tao Qin
3. Tie-Yan Liu
4. Yu Zhang
5. Frank K. Soong
6. Hung-yi Lee
7. Ye Jia
8. Yonghui Wu

Citations in the corpus

• Transfer Learning from Speaker Verification to Multispeaker Text-To-Speech Synthesis (2018, methodology): “[7] used a similar transfer learning configuration where a speaker embedding computed from a pre-trained speaker classifier was used to condition a TTS system.”
• A Survey on Neural Speech Synthesis (2021, background): “, a lot of works in statistic parametric speech synthesis have studied voice adaptation [77, 386, 443, 65, 122], and the recent voice cloning challenge also attracts a lot of participants [388, 118, 331, 45].”
• NAUTILUS: A Versatile Voice Cloning System (2020, background): “d-vector [18], [48], or x-vector [49], which are all byproducts of speaker recognition systems.”
• Neural Text-to-Speech Adaptation from Low Quality Public Recordings (2019, background): “…onehot lookup table embedding no sv speaker verification output no onehot-pf lookup table embedding yes sv-pf speaker verification system yes noid-pf none yes tinguish the speaker, [16, 18, 27] have introduced a trainable speaker encoder to map the speaker’s voice into a latent speaker space.”
• Speaker Adaptation of a Multilingual Acoustic Model for Cross-Language Synthesis (2020, background): “Second, the input to the acoustic model comprises not only linguistic features (including the language code) but also speaker embed-dings extracted from the neural networks (e.g., d-vectors) [27, 9].”
• Zero-Shot Text-to-Speech Synthesis Conditioned Using Self-Supervised Speech Representation Model (2023, methodology): “Two main approaches have been used to obtain such embeddings: speaker-recognition-based embedding such as d-vector [4, 5] and x-vector [3, 6], and embedding from the reference encoder trained simultaneously with the acoustic model, such as using global style tokens (GSTs) [7, 8].”
• SpeechNet: A Universal Modularized Model for Speech Processing Tasks (2021, methodology): “, 2018), extracted with an off-the-shelf speaker embedding model (Doddipatla et al., 2017), or encoded by Speaker Encoder from an input speech X.”