Voice Acting Pipeline Output: Synthetic Emotional Speech with Perceptual Scores

Voice Acting Pipeline Output is a synthetic emotional speech dataset generated by an automated, multi-GPU system. Each sample consists of 6 audio generations from a consistent speaker, scored across 59 perceptual dimensions by Empathic Insight Voice+. The dataset was created by TTS-AGI and was last updated on March 31, 2026.

Use Cases

• Training emotional TTS models based on controlled, disentangled speaker and emotion data.
• Benchmarking speech synthesis quality using the 59 perceptual dimensions provided.
• Studying the relationship between speaker identity and emotional prosody in synthetic speech.

Strengths

• Each sample includes 6 audio generations (3 emotional and 3 neutral sentences) from a consistent speaker.
• Samples are scored across 59 perceptual dimensions (55 for emotion/attributes and 4 for quality).
• Data is generated by an automated, multi-GPU pipeline designed for controlled TTS training.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Data may reflect bias inherent to the synthetic generation process and source platform.

Provenance

Source

TTS-AGI via Hugging Face.

Collection Method

Generated by the Voice Acting Pipeline, an automated, multi-GPU data generation system.

Freshness

Last updated 2026-03-31 13:04:45; freshness should be verified.