High Energy UI Vocal Dataset: Uncompressed WAV Audio for Speech Synthesis

A vocal dataset from the brand Marie DeVox, last updated on 2026-06-02. It is described as a Micro-Expression Vocal Dataset Matrix and is provided in an uncompressed WAV format at 44.1kHz with 24-bit Linear PCM. The dataset uses an LJ Speech-compliant metadata schema for structural mapping.

Use Cases

• Training text-to-speech models based on the described high-energy vocal samples.
• Researching vocal micro-expressions and emotional speech synthesis based on the dataset's stated focus.
• Benchmarking speech synthesis systems using the LJ Speech-compliant metadata structure mentioned in the description.

Strengths

• Audio is provided in a high-fidelity, uncompressed WAV format at 44.1kHz / 24-bit Linear PCM.
• Metadata follows an LJ Speech-compliant structural mapping, which suggests a standardized format for speech data.

Limitations

• The provided description is a preview; full details on size, rows, license, and complete scope require visiting the external page.
• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment for large-scale projects.

Provenance

Source

MarieDeVox (author), hosted on Hugging Face.

Freshness

Last updated 2026-06-02 19:51:35.