Speaker Adaptive Training using Model Agnostic Meta-Learning

2019-10-23 · Ondřej Klejch, Joachim Fainberg, Peter Bell, Steve Renals

Speaker adaptive training (SAT) of neural network acoustic models learns models in a way that makes them more suitable for adaptation to test conditions. Conventionally, model-based speaker adaptive training is performed by having a set of speaker dependent parameters that are jointly optimised with speaker independent parameters in order to remove speaker variation. However, this does not scale well if all neural network weights are to be adapted to the speaker. In this paper we formulate speaker adaptive training as a meta-learning task, in which an adaptation process using gradient descent is encoded directly into the training of the model. We compare our approach with test-only adaptation of a standard baseline model and a SAT-LHUC model with a learned speaker adaptation schedule and demonstrate that the meta-learning approach achieves comparable results.

📄 PDF Abstract BibTeX arXiv:1910.10605

Code (1)

ondrejklejch/learning_to_adapt 공식 구현 tf

Tasks

Meta-Learningmodel

Methods 이 논문이 사용한 방법론

Test 설명 없음

Meta-TTS: Meta-Learning for Few-Shot Speaker Adaptive Text-to-Speech

2021-11-07 · Sung-Feng Huang, Chyi-Jiunn Lin, Da-Rong Liu, Yi-Chen Chen 외

Personalizing a speech synthesis system is a highly desired application, where the system can generate speech with the user's voice with rare enrolled recordings. There are two main approaches to build such a system in r…

Meta-Learning Speech Synthesis text-to-speech Text to Speech

Meta-Voice: Fast few-shot style transfer for expressive voice cloning using meta learning

2021-11-14 · Songxiang Liu, Dan Su, Dong Yu

The task of few-shot style transfer for voice cloning in text-to-speech (TTS) synthesis aims at transferring speaking styles of an arbitrary source speaker to a target speaker's voice using very limited amount of neutral…

Disentanglement Meta-Learning Style Transfer text-to-speech+2

Sample Efficient Adaptive Text-to-Speech

2018-09-27 · ICLR 2019 5 · Yutian Chen, Yannis Assael, Brendan Shillingford, David Budden 외

We present a meta-learning approach for adaptive text-to-speech (TTS) with few data. During training, we learn a multi-speaker model using a shared conditional WaveNet core and independent learned embeddings for each spe…

Meta-Learning text-to-speech Text to Speech Voice Similarity

Improved Meta-Learning Training for Speaker Verification

2021-03-29 · Yafeng Chen, Wu Guo, Bin Gu

Meta-learning has recently become a research hotspot in speaker verification (SV). We introduce two methods to improve the meta-learning training for SV in this paper. For the first method, a backbone embedding network i…

Data Augmentation Meta-Learning Speaker Verification speech-recognition+1

Meta-StyleSpeech : Multi-Speaker Adaptive Text-to-Speech Generation

2021-06-06 · Dongchan Min, Dong Bok Lee, Eunho Yang, Sung Ju Hwang

With rapid progress in neural text-to-speech (TTS) models, personalized speech generation is now in high demand for many applications. For practical applicability, a TTS model should generate high-quality speech with onl…

text-to-speech Text to Speech