# S-DiverSe

**S-DiverSe: Spanish Diverse Speech**

Authors: Fernando López, Fernando Ibañez, Ana Martínez, Iván Alonso, Pablo Gómez, Santosh Kesiraju, Jordi Luque

## About

Automatic speech recognition (ASR) has advanced remarkably for standard speech, yet speech affected by neurological conditions remains a challenge. We present S-DiverSe (Spanish Diverse Speech), a corpus of 3.2 hours of in-the-wild Spanish speech from 22 speakers with amyotrophic lateral sclerosis, Parkinson's disease, and stroke. The dataset contains 444 manually transcribed audio segments with metadata on speaker sex, disease type, and intelligibility. S-DiverSe is designed to support ASR evaluation and development for neurologically affected Spanish speech. We describe the dataset, analyze its composition, and report baseline ASR results alongside initial adaptation experiments. Our findings reveal that heuristic text post-processing is more robust than fine-tuning for out-of-domain neurological Spanish speech. This underscores the need for dedicated in-the-wild Spanish benchmarks.


## Setup environment

Install dependencies:

```bash
python3 -m venv .venv
source .venv/bin/activate
pip install yt-dlp
```

You also need [Deno](https://deno.land/) for yt-dlp to work properly with YouTube:

```bash
curl -fsSL https://deno.land/install.sh | sh
```

After installing, restart your terminal or run:

```bash
export PATH="$HOME/.deno/bin:$PATH"
```

## Download audio

The script reads a TSV file with YouTube links and timestamps, downloads the audio, and extracts the segments.

```bash
mkdir -p data/audio
python src/download_audio.py --tsv_file data/diversse.tsv
```

Output files are saved to `data/audio/` organized by disease folder (e.g., `data/audio/ALS/`).

## Citation

If you use this corpus in your research, please cite:

```bibtex
@misc{lópez2026sdiverse
      title={S-DiverSe: Spanish Diverse Speech},
      author={Fernando López and Fernando Ibañez and Ana Martínez and Iván Alonso and Pablo Gómez and Santosh Kesiraju and Jordi Luque},
      year={2026},
      eprint={2607.03207},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2607.03207},
}
```