Datasetpublic

Firoj112/maithili_syspin_female_tts_22050

Maithili TTS Dataset (IISc SYSPIN Female) This is a Maithili female TTS dataset from the IISc SYSPIN project. It has been converted to 22050 Hz (mono) for seamless use in TTS fine-tuning, following the same schema as Firoj112/nepali_openslr43_tts_22050. Dataset Summary Language: Maithili (mai) Speaker: Spk0001 (Female) Total Duration: ~59 hours 40 mins Total Utterances: 34,412 Sampling Rate: 22050 Hz (Resampled from 48kHz) Format: Mono channel, float32 PCM… See the full description on the dataset page: https://huggingface.co/datasets/Firoj112/maithili_syspin_female_tts_22050.

sourceHugging Facecc-by-4.0updated 6mo agoView on Hugging Face

0likes77downloads

Dataset Card

Maithili TTS Dataset (IISc SYSPIN Female)

This is a Maithili female TTS dataset from the IISc SYSPIN project. It has been converted to 22050 Hz (mono) for seamless use in TTS fine-tuning, following the same schema as Firoj112/nepali_openslr43_tts_22050.

Dataset Summary

—Language: Maithili (mai)
—Speaker: Spk0001 (Female)
—Total Duration: ~59 hours 40 mins
—Total Utterances: 34,412
—Sampling Rate: 22050 Hz (Resampled from 48kHz)
—Format: Mono channel, float32 PCM

Dataset Schema

Each entry contains:

—audio: Audio feature at 22050 Hz (resampled).
—text: The transcript string.
—speaker_id: String identifying the speaker ("Spk0001").
—language: ISO code for Maithili ("mai").
—id: Formatted unique ID (mai_Spk0001_{original_id}).

Domain Breakdown

The dataset covers 12 domains from the original SYSPIN corpus:

—AGRICULTURE (AGRI)
—BOOKS (BOOK)
—EVALUATION (EVAL)
—FINANCE (FINA)
—FOOD (FOOD)
—HEALTH (HEAL)
—INDIA RELATED (INDI)
—LOCAL CONVERSATION (LOCA)
—POLITICS (POLI)
—SOCIAL (SOCI)
—SPORTS (SPOR)
—TECHNOLOGY (TECH)

Source & Attribution

Original dataset collected by Indian Institute of Science (IISc), Bengaluru as part of the SYSPIN project (2024). Project Link: spiredatasets.iisc.ac.in/syspinCorpus

Citation

bibtex

@misc{SYSPIN_S1.0_Corpus,
    Title = {SYSPIN_S1.0 Corpus - A TTS Corpus of 900+ hours in nine Indian Languages},
    Authors = {Abhayjeet Et al.},
    Year = {2025}
}

License

CC BY 4.0

Prepared by: Firoj112