OVSpeech: Open-Vocabulary Instruct Text-to-Speech Dataset

OVSpeech is a dataset built for the ICASSP 2026 paper titled 'OV-InstructTTS: Towards Open-Vocabulary Instruct Text-to-Speech'. It is constructed upon the ContextSpeech framework and is authored by y-ren16. The dataset was last updated on the Hugging Face platform in April 2026.

Use Cases

• Training open-vocabulary text-to-speech models based on the described ContextSpeech framework.
• Benchmarking instruction-following capabilities in speech synthesis systems.
• Researching methods to expand the vocabulary and control parameters of neural TTS models.

Strengths

• Dataset is associated with a peer-reviewed ICASSP 2026 conference paper.
• Built upon an established framework, ContextSpeech, suggesting a structured foundation.

Limitations

• Description metadata is limited; actual data quality requires manual inspection after download.
• Column-level documentation is absent; field semantics must be inferred after download.
• Row count and file size are unknown, which may limit suitability assessment.

Provenance

Source

Hugging Face platform, author y-ren16.

Collection Method

Built upon the ContextSpeech framework; specific collection method not detailed.

Time Range

Associated with research for ICASSP 2026.

Freshness

Last updated 2026-04-22 11:36:37; freshness should be verified.

Geography

null