Text-to-speech Synthesis In The Wild

Abstract

Traditional Text-to-Speech (TTS) systems rely on studio-quality speech recorded in controlled settings.a Recently, an effort known as noisy-TTS training has emerged, aiming to utilize in-the-wild data. However, the lack of dedicated datasets has been a significant limitation. We introduce the TTS In the Wild (TITW) dataset, which is publicly available, created through a fully automated pipeline applied to the VoxCeleb1 dataset. It comprises two training sets: TITW-Hard, derived from the transcription, segmentation, and selection of raw VoxCeleb1 data, and TITW-Easy, which incorporates additional enhancement and data selection based on DNSMOS. State-of-the-art TTS models achieve over 3.0 UTMOS score with TITW-Easy, while TITW-Hard remains difficult showing UTMOS below 2.8.

Stats

• citations0
• S2 citations—
• github stars0
• HF likes0
• heat score0.00
• arxiv keyjung2024text

Related papers

• Wenetspeech4tts: A 12,800-hour Mandarin TTS Corpus For Large Speech Generation Model Benchmark (2024)9.76
• The Zero Resource Speech Challenge 2019: TTS Without T (2019)13.17
• An Automated End-to-end Open-source Software For High-quality Text-to-speech Dataset Generation (2024)0.00
• Mntts2: An Open-source Multi-speaker Mongolian Text-to-speech Synthesis Dataset (2022)5.81
• More Than Words: In-the-wild Visually-driven Prosody For Text-to-speech (2021)9.03
• BASE TTS: Lessons From Building A Billion-parameter Text-to-speech Model On 100K Hours Of Data (2024)0.00
• Libri2vox Dataset: Target Speaker Extraction With Diverse Speaker Conditions And Synthetic Data (2024)0.00
• Mntts: An Open-source Mongolian Text-to-speech Synthesis Dataset And Accompanied Baseline (2022)5.24