Datasets:
pretty_name: Urdu Speech Samples
license: other
license_name: poseidon-sample-preview
license_link: https://huggingface.co/datasets/psdn-ai/urdu-speech-samples/blob/main/TERMS.md
tags:
- audio
- urdu
- speech
- transcripts
- samples
language:
- ur
task_categories:
- automatic-speech-recognition
size_categories:
- n<1K
Urdu Speech Samples
This sample shows Urdu speech with transcript alignment and simple audio metadata. It is meant to help buyers review language fit and capture quality before requesting a larger sample or production delivery.
What This Shows
- Urdu speech audio with paired text
- A compact view of transcript and metadata structure
- Audio format signals for procurement review
Dataset Specifications
| Field | Value |
|---|---|
| Modality | Audio |
| Language | Urdu |
| Audio format | Audio clips |
| Sample rate | 48 kHz |
| Channels | Mono |
| Transcript source | Ground truth |
| Sample size | 1 clip |
Metadata Coverage
The value is not just the audio clip; it is the combination of listenable media, transcript status, language labels, speaker context, and technical audio metadata that lets a buyer evaluate fit before requesting broader coverage.
Sample metadata includes: sample identifiers, sample titles, language labels, language code, duration, media filename, codec, sample rate, channels, file size, transcript text, transcript source, PII status, tags, feature flags, acoustic scene, codec detail, emotion, intent, sentiment, and audio type.
Sample Scope
The sample currently includes 1 clip. Ground-truth transcript metadata is included where available; we can add more review coverage when needed. Larger deliveries can be scoped by coverage, format, volume, metadata depth, and licensing requirements.
For broader coverage or custom requirements, contact hello@psdn.ai or visit psdn.ai/contact.