[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-82719-en":3,"doc-seo-82719-105":29,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":21,"is_downloadable":21,"audit_status":21,"page_count":20,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":13,"seo_description":14,"update_tm":27,"read_time":28},82719,4810365810221,"Aurora","https://ap-avatar.wpscdn.com/davatar_155a257f0dc6eb9ab79c44ca47cae57d",8,"Research & Report","S-DiverSe Spanish Diverse Speech Corpus and Baseline ASR Results","Automatic speech recognition has made major progress on standard speech, but neurologically affected speech remains difficult. S-DiverSe (Spanish Diverse Speech) provides a 3.2-hour in-the-wild Spanish corpus from 22 speakers diagnosed with amyotrophic lateral sclerosis, Parkinson’s disease, and stroke. The dataset includes 444 manually transcribed audio segments with metadata on speaker sex, disease type, and intelligibility. The work analyzes corpus composition, reports baseline ASR and adaptation experiments, and shows heuristic post-processing is more robust than fine-tuning for out-of-domain neurological speech.","S-DiverSe: Spanish Diverse Speech  \nFernando Lo´pez  1 ,2 ,∗ ,∗∗, Fernando Iban˜ez2 ,∗, Ana Martı´nez2, Iva´n Alonso2,  \nPablo Go´mez 1, Santosh Kesiraju  3, Jordi Luque  1  \n1 Scientific Research, Telefnica Innovacin Digital, Spain  \n2 Universidad Autnoma de Madrid, Spain  \n3 Brno University of Technology, Czech Republic  \n[fernando.lopez@telefonica.com](fernando.lopez@telefonica.com)  \narXiv :2607 .03207v 1 [ cs .CL] 3 Jul 2026  \nAbstract  \nAutomatic 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-ofdomain neurological Spanish speech. This underscores the need for dedicated in-the-wild Spanish benchmarks.  \nIndex Terms: speech recognition, pathological speech, speech corpus, neurological disorders  \n1. Introduction  \nAutomatic speech recognition (ASR) has improved markedly in recent years, with state-of-the-art models achieving low word error rates on standard benchmarks [1] . However, they continue to face substantial challenges when applied to real-world speech scenarios [2] . One of these challenges is the recognition of speech produced by individuals with neuromotor disorders, such as amyotrophic lateral sclerosis (ALS), Parkinson’s disease (PD), and post-stroke conditions. These conditions frequently cause dysarthria, which impairs neuromuscular control of speech and leads to reduced articulation clarity, altered prosody, and variable intelligibility, all of which challenge ASR systems [3, 4] .  \nA central obstacle in pathological speech recognition is the lack of large, high-quality datasets [5] . This remains true even for English, despite resources such as UA-Speech [4], TORGO [3], and the recent Interspeech Speech Accessibility Project (SAP) challenge [6] . For Spanish, the gap is wider: publicly available corpora for neurological conditions are scarce and typically limited in either clinical coverage or domain diversity.  \nThe Chilean Spanish dataset from the GITA laboratory [7] is a significant early contribution. It encompasses phonation tasks such as isolated vowels, vowel sequences, and changing tones, alongside diadochokinetic evaluations, word and sentence repetitions, and spontaneous speech. However, it lacks diversity in recording conditions and is predominantly biased towards an elderly demographic. More recently, NeuroVoz [8, 9]  \n*These authors contributed equally.  \n**indicates the corresponding author.  \nhas provided Castilian Spanish data for PD collected under controlled hospital conditions, including sustained vowels, sentence repetitions, diadochokinetic evaluation, and short monologues. While valuable, it primarily contains brief, elicited utterances from elderly participants, limiting its ability to capture the broader variability of neurologically affected speech.  \nConsequently, rigorous evaluation of Spanish ASR for speech affected by different neurological conditions remains difficult in the absence of benchmarks. This limitation hinders both the development of robust ASR for pathological speech and progress on assistive communication and clinical assessment tools. To bridge this gap, we make three contributions:(i) We introduce S-DiverSe (Spanish Diverse Speech), the first in-the-wild Spanish dataset covering multiple neurological conditions. It includes human transcripts and ","cbCaic1KGT0CwDcu","https://ap.wps.com/l/cbCaic1KGT0CwDcu","pdf",315597,5,1,"English","en",105,"# Introduction\n# S-DiverSe: Spanish Diverse Speech\n## Statistics","[{\"question\":\"What is S-DiverSe and who does it include?\",\"answer\":\"S-DiverSe is a Spanish in-the-wild speech corpus. It contains recordings from 22 speakers with amyotrophic lateral sclerosis, Parkinson’s disease, and stroke, and it provides human transcripts with metadata.\"},{\"question\":\"How is the dataset organized and what metadata does it provide?\",\"answer\":\"The corpus contains 3.2 hours of speech and 444 manually transcribed audio segments. It includes metadata for speaker sex, pathological condition, and intelligibility.\"},{\"question\":\"What adaptation approaches are evaluated and what is the main finding?\",\"answer\":\"Four state-of-the-art ASR systems are evaluated, and post-processing and fine-tuning techniques are tested for two ASR models. Heuristic text post-processing is more robust than fine-tuning for out-of-domain neurological Spanish speech.\"}]",1784182473,13,{"code":4,"msg":30,"data":31},"ok",{"site_id":24,"language":23,"slug":32,"title":13,"keywords":33,"description":14,"schema_data":34,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":27},"s-diverse-spanish-diverse-speech-corpus-and-baseline-asr-results","",{"@graph":35,"@context":85},[36,53,68],{"@type":37,"itemListElement":38},"BreadcrumbList",[39,43,47,50],{"item":40,"name":41,"@type":42,"position":21},"https://docshare.wps.com","Home","ListItem",{"item":44,"name":45,"@type":42,"position":46},"https://docshare.wps.com/document/","Document",2,{"item":48,"name":12,"@type":42,"position":49},"https://docshare.wps.com/document/research-report/",3,{"item":51,"name":13,"@type":42,"position":52},"https://docshare.wps.com/document/s-diverse-spanish-diverse-speech-corpus-and-baseline-asr-results/82719/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":23,"description":14,"dateModified":61,"datePublished":62,"encodingFormat":60,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":40,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-24","2026-07-16",true,{"@type":65,"interactionType":66,"userInteractionCount":20},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"What is S-DiverSe and who does it include?","Question",{"text":75,"@type":76},"S-DiverSe is a Spanish in-the-wild speech corpus. It contains recordings from 22 speakers with amyotrophic lateral sclerosis, Parkinson’s disease, and stroke, and it provides human transcripts with metadata.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How is the dataset organized and what metadata does it provide?",{"text":80,"@type":76},"The corpus contains 3.2 hours of speech and 444 manually transcribed audio segments. It includes metadata for speaker sex, pathological condition, and intelligibility.",{"name":82,"@type":73,"acceptedAnswer":83},"What adaptation approaches are evaluated and what is the main finding?",{"text":84,"@type":76},"Four state-of-the-art ASR systems are evaluated, and post-processing and fine-tuning techniques are tested for two ASR models. Heuristic text post-processing is more robust than fine-tuning for out-of-domain neurological Spanish speech.","https://schema.org",{"og:url":51,"og:type":87,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":89,"canonical":51},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,109,114,119,122,127,130,134],{"id":21,"doc_module":4,"doc_module_name":45,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":46,"doc_module":4,"doc_module_name":45,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":45,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":20,"doc_module":4,"doc_module_name":45,"category_name":106,"show_sort_weight":107,"slug":108},"Comic",60,"comic",{"id":110,"doc_module":4,"doc_module_name":45,"category_name":111,"show_sort_weight":112,"slug":113},6,"Technology",50,"technology",{"id":115,"doc_module":4,"doc_module_name":45,"category_name":116,"show_sort_weight":117,"slug":118},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":120,"slug":121},30,"research-report",{"id":123,"doc_module":4,"doc_module_name":45,"category_name":124,"show_sort_weight":125,"slug":126},9,"Religion & Spirituality",20,"religion-spirituality",{"id":125,"doc_module":4,"doc_module_name":45,"category_name":128,"show_sort_weight":125,"slug":129},"World Cup","world-cup",{"id":131,"doc_module":4,"doc_module_name":45,"category_name":132,"show_sort_weight":131,"slug":133},10,"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":45,"category_name":136,"show_sort_weight":20,"slug":137},19,"General","general"]