Generating Synthetic Audio Data for Attention-Based Speech Recognition Systems
By Nick Rossenbach, Albert Zeyer, Ralf Schlüter, Hermann Ney. 2019
96 citations recorded by the source corpus.
Recent advances in text-to-speech (TTS) led to the development of flexible multi-speaker end-to-end TTS systems. We extend state-of-the-art attention-based automatic speech recognition (ASR) systems with synthetic audio generated by a TTS system trained only on the ASR corpora itself. ASR and TTS systems are built separately to show that text-only data can be used to enhance existing end-to-end ASR systems without the necessity of parameter or architecture changes. We compare our method with language model integration of the same text data and with simple data augmentation methods like SpecAugment and show that performance improvements are mostly independent. We achieve improvements of up to 33% relative in word-error-rate (WER) over a strong baseline with data-augmentation in a low-resource environment (LibriSpeech-100h), closing the gap to a comparable oracle experiment by more than 50%. We also show improvements of up to 5% relative WER over our most recent ASR baseline on LibriSpeech-960h.
Subfields: Generative Models, Speech & Audio
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
- Andrei Andrusenko
- Shinsuke Sakai
- Aleksandr Laptev
- Tatsuya Kawahara
- Ivan Medennikov
- Masato Mimura
- Sei Ueno
- Jinyu Li
Citations in the corpus
- You Do Not Need More Data: Improving End-To-End Speech Recognition by Text-To-Speech Data Augmentation (2020, positive): “In contrast, in this work we follow [13] and use LibriSpeech database [14] to simulate a so-called low-to-medium-resource (here stands for one hundred hours) task.”
- Exploring Machine Speech Chain for Domain Adaptation and Few-Shot Speaker Adaptation (2021, unknown): “In [23], TTS and ASR are trained separately, which can avoid the effort of training new models using the same acoustic feature for joint training.”
- Data Augmentation for ASR Using TTS Via a Discrete Representation (2021, background): “Several studies have used speaker information to generate multi-speaker speech such as a speaker ID [14], VAE latent variables [15], pre-trained speaker verification model [21], and a global style token (GST) [16].”