Datasets:

---

MERaLiON

/

sea_audiobench_datasets_SpR

Like 0

Follow

MERaLiON 125

Adds a dataset card for the SpkR (speaker recognition) task of SEA-SpeechBench.

Contents:

• YAML metadata: license, languages, task category, and gated-access notice for the non-commercial source material
• Loading example for datasets>=4.0
• Subset table: 9 configs, 8,827 examples, source corpus and audio duration per config
• Data field descriptions for the paired-context schema
• Source data table with upstream licenses
• BibTeX for the benchmark and all eight source corpora