# The Zero Resource Speech Challenge 2019: TTS without T

## TL;DR

The Zero Resource Speech Challenge 2019 aims to develop speech synthesis systems without relying on text or phonetic labels, using only raw audio data and unsupervised learning to discover subword units for TTS.

## Contribution

This paper introduces a novel challenge for unsupervised speech synthesis, providing datasets, evaluation metrics, and baseline systems to foster research in TTS without text.

## Key findings

- 19 systems submitted by 10 teams demonstrate diverse approaches.
- Unsupervised subword discovery can produce usable units for TTS.
- Baseline systems show the potential and challenges of zero-resource TTS.

## Abstract

We present the Zero Resource Speech Challenge 2019, which proposes to build a speech synthesizer without any text or phonetic labels: hence, TTS without T (text-to-speech without text). We provide raw audio for a target voice in an unknown language (the Voice dataset), but no alignment, text or labels. Participants must discover subword units in an unsupervised way (using the Unit Discovery dataset) and align them to the voice recordings in a way that works best for the purpose of synthesizing novel utterances from novel speakers, similar to the target speaker's voice. We describe the metrics used for evaluation, a baseline system consisting of unsupervised subword unit discovery plus a standard TTS system, and a topline TTS using gold phoneme transcriptions. We present an overview of the 19 submitted systems from 10 teams and discuss the main results.