Neyshekar V5: Persian Speech Recordings for ASR and TTS

Neyshekar is an open, community-driven Persian speech dataset collected via a web-based crowdsourcing platform. Recordings are provided by volunteer contributors and paid native Persian voice actors. The audio is mono, 16-bit PCM, sampled at 16 kHz.

Use Cases

• Train Persian automatic speech recognition models based on native speaker recordings.
• Develop text-to-speech systems based on the crowdsourced audio samples.
• Perform speech representation learning for downstream Persian language tasks.
• Fine-tune speech models for specific Persian dialects or accents based on the contributor pool.

Strengths

• Audio recordings are mono, 16-bit PCM, sampled at 16 kHz, which is a standard format for speech processing.
• Data is collected from native Persian speakers, including both volunteers and paid voice actors, which may improve quality and diversity.

Limitations

• Row count is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• Last updated 2026-07-16 15:52:05; freshness should be verified.

Provenance

Source

shekar-ai via Hugging Face

Collection Method

Collected via a web-based crowdsourcing platform at ney.shekar.io from volunteer contributors and paid voice actors.

Freshness

Last updated 2026-07-16 15:52:05.

Geography

Persian language speakers; specific geography is not stated.