Improving Child Speech Recognition and Reading Mistake Detection by Using Prompts Article Swipe

View

Lingyun Gao , Cristian Tejedor-Garcı́a , Catia Cucchiarini , Helmer Strik ·

YOU? · · 2025 · Open Access · · DOI: https://doi.org/10.21437/interspeech.2025-658

Automatic reading aloud evaluation can provide valuable support to teachers by enabling more efficient scoring of reading exercises. However, research on reading evaluation systems and applications remains limited. We present a novel multimodal approach that leverages audio and knowledge from text resources. In particular, we explored the potential of using Whisper and instruction-tuned large language models (LLMs) with prompts to improve transcriptions for child speech recognition, as well as their effectiveness in downstream reading mistake detection. Our results demonstrate the effectiveness of prompting Whisper and prompting LLM, compared to the baseline Whisper model without prompting. The best performing system achieved state-of-the-art recognition performance in Dutch child read speech, with a word error rate (WER) of 5.1%, improving the baseline WER of 9.4%. Furthermore, it significantly improved reading mistake detection, increasing the F1 score from 0.39 to 0.73.

Related Topics

Rivers Of Blood Speech

Julia Child

Child Sexual Abuse

Monday's Child

Child's Play (Franchise)

The King's Speech

Feral Child

Child's Play (1988 Film)

Desmond Child

George S. Patton's Speech To The Third Army

Child Erotica

Child's Play 2

Concepts

No concepts available.

Metadata

Type: article
Language: en
Landing Page: https://doi.org/10.21437/interspeech.2025-658
OA Status: green
OpenAlex ID: https://openalex.org/W4415433901

All OpenAlex metadata

Raw OpenAlex JSON

OpenAlex ID: https://openalex.org/W4415433901

Canonical identifier for this work in OpenAlex
DOI: https://doi.org/10.21437/interspeech.2025-658

Digital Object Identifier
Title: Improving Child Speech Recognition and Reading Mistake Detection by Using Prompts

Work title
Type: article

OpenAlex work type
Language: en

Primary language
Publication year: 2025

Year of publication
Publication date: 2025-08-17

Full publication date if available
Authors: Lingyun Gao, Cristian Tejedor-Garcı́a, Catia Cucchiarini, Helmer Strik

List of authors in order
Landing page: https://doi.org/10.21437/interspeech.2025-658

Publisher landing page
Open access: Yes

Whether a free full text is available
OA status: green

Open access status per OpenAlex
OA URL: https://arxiv.org/pdf/2506.11079

Direct OA link when available
Cited by: 0

Total citation count in OpenAlex

Full payload

id	https://openalex.org/W4415433901
doi	https://doi.org/10.21437/interspeech.2025-658
ids.doi	https://doi.org/10.21437/interspeech.2025-658
ids.openalex	https://openalex.org/W4415433901
fwci	0.0
type	article
title	Improving Child Speech Recognition and Reading Mistake Detection by Using Prompts
biblio.issue
biblio.volume
biblio.last_page	2854
biblio.first_page	2850
topics[0].id	https://openalex.org/T10201
topics[0].field.id	https://openalex.org/fields/17
topics[0].field.display_name	Computer Science
topics[0].score	0.9021000266075134
topics[0].domain.id	https://openalex.org/domains/3
topics[0].domain.display_name	Physical Sciences
topics[0].subfield.id	https://openalex.org/subfields/1702
topics[0].subfield.display_name	Artificial Intelligence
topics[0].display_name	Speech Recognition and Synthesis
is_xpac	False
apc_list
apc_paid
language	en
locations[0].id	doi:10.21437/interspeech.2025-658
locations[0].is_oa	False
locations[0].source
locations[0].license
locations[0].pdf_url
locations[0].version	publishedVersion
locations[0].raw_type	proceedings-article
locations[0].license_id
locations[0].is_accepted	True
locations[0].is_published	True
locations[0].raw_source_name	Interspeech 2025
locations[0].landing_page_url	https://doi.org/10.21437/interspeech.2025-658
locations[1].id	pmh:oai:arXiv.org:2506.11079
locations[1].is_oa	True
locations[1].source.id	https://openalex.org/S4306400194
locations[1].source.issn
locations[1].source.type	repository
locations[1].source.is_oa	True
locations[1].source.issn_l
locations[1].source.is_core	False
locations[1].source.is_in_doaj	False
locations[1].source.display_name	arXiv (Cornell University)
locations[1].source.host_organization	https://openalex.org/I205783295
locations[1].source.host_organization_name	Cornell University
locations[1].source.host_organization_lineage	https://openalex.org/I205783295
locations[1].license
locations[1].pdf_url	https://arxiv.org/pdf/2506.11079
locations[1].version	submittedVersion
locations[1].raw_type	text
locations[1].license_id
locations[1].is_accepted	False
locations[1].is_published	False
locations[1].raw_source_name
locations[1].landing_page_url	http://arxiv.org/abs/2506.11079
indexed_in	arxiv, crossref
authorships[0].author.id	https://openalex.org/A5079316816
authorships[0].author.orcid	https://orcid.org/0000-0003-2509-9505
authorships[0].author.display_name	Lingyun Gao
authorships[0].author_position	first
authorships[0].raw_author_name	Lingyun Gao
authorships[0].is_corresponding	False
authorships[1].author.id	https://openalex.org/A5046936545
authorships[1].author.orcid	https://orcid.org/0000-0001-5395-0438
authorships[1].author.display_name	Cristian Tejedor-Garcı́a
authorships[1].author_position	middle
authorships[1].raw_author_name	Cristian Tejedor-Garcia
authorships[1].is_corresponding	False
authorships[2].author.id	https://openalex.org/A5006678637
authorships[2].author.orcid	https://orcid.org/0000-0001-5908-0824
authorships[2].author.display_name	Catia Cucchiarini
authorships[2].author_position	middle
authorships[2].raw_author_name	Catia Cucchiarini
authorships[2].is_corresponding	False
authorships[3].author.id	https://openalex.org/A5019585114
authorships[3].author.orcid	https://orcid.org/0000-0003-1722-3465
authorships[3].author.display_name	Helmer Strik
authorships[3].author_position	last
authorships[3].raw_author_name	Helmer Strik
authorships[3].is_corresponding	False
has_content.pdf	False
has_content.grobid_xml	False
is_paratext	False
open_access.is_oa	True
open_access.oa_url	https://arxiv.org/pdf/2506.11079
open_access.oa_status	green
open_access.any_repository_has_fulltext	False
created_date	2025-10-23T00:00:00
display_name	Improving Child Speech Recognition and Reading Mistake Detection by Using Prompts
has_fulltext	False
is_retracted	False
updated_date	2025-11-06T03:46:38.306776
primary_topic.id	https://openalex.org/T10201
primary_topic.field.id	https://openalex.org/fields/17
primary_topic.field.display_name	Computer Science
primary_topic.score	0.9021000266075134
primary_topic.domain.id	https://openalex.org/domains/3
primary_topic.domain.display_name	Physical Sciences
primary_topic.subfield.id	https://openalex.org/subfields/1702
primary_topic.subfield.display_name	Artificial Intelligence
primary_topic.display_name	Speech Recognition and Synthesis
cited_by_count	0
locations_count	2
best_oa_location.id	pmh:oai:arXiv.org:2506.11079
best_oa_location.is_oa	True
best_oa_location.source.id	https://openalex.org/S4306400194
best_oa_location.source.issn
best_oa_location.source.type	repository
best_oa_location.source.is_oa	True
best_oa_location.source.issn_l
best_oa_location.source.is_core	False
best_oa_location.source.is_in_doaj	False
best_oa_location.source.display_name	arXiv (Cornell University)
best_oa_location.source.host_organization	https://openalex.org/I205783295
best_oa_location.source.host_organization_name	Cornell University
best_oa_location.source.host_organization_lineage	https://openalex.org/I205783295
best_oa_location.license
best_oa_location.pdf_url	https://arxiv.org/pdf/2506.11079
best_oa_location.version	submittedVersion
best_oa_location.raw_type	text
best_oa_location.license_id
best_oa_location.is_accepted	False
best_oa_location.is_published	False
best_oa_location.raw_source_name
best_oa_location.landing_page_url	http://arxiv.org/abs/2506.11079
primary_location.id	doi:10.21437/interspeech.2025-658
primary_location.is_oa	False
primary_location.source
primary_location.license
primary_location.pdf_url
primary_location.version	publishedVersion
primary_location.raw_type	proceedings-article
primary_location.license_id
primary_location.is_accepted	True
primary_location.is_published	True
primary_location.raw_source_name	Interspeech 2025
primary_location.landing_page_url	https://doi.org/10.21437/interspeech.2025-658
publication_date	2025-08-17
publication_year	2025
referenced_works_count	0
abstract_inverted_index.a	30, 109
abstract_inverted_index.F1	131
abstract_inverted_index.In	42
abstract_inverted_index.We	28
abstract_inverted_index.as	66, 68
abstract_inverted_index.by	10
abstract_inverted_index.in	71, 103
abstract_inverted_index.it	123
abstract_inverted_index.of	15, 48, 81, 114, 120
abstract_inverted_index.on	20
abstract_inverted_index.to	8, 59, 88, 135
abstract_inverted_index.we	44
abstract_inverted_index.Our	76
abstract_inverted_index.The	95
abstract_inverted_index.WER	119
abstract_inverted_index.and	24, 37, 51, 84
abstract_inverted_index.can	4
abstract_inverted_index.for	62
abstract_inverted_index.the	46, 79, 89, 117, 130
abstract_inverted_index.0.39	134
abstract_inverted_index.LLM,	86
abstract_inverted_index.best	96
abstract_inverted_index.from	39, 133
abstract_inverted_index.more	12
abstract_inverted_index.rate	112
abstract_inverted_index.read	106
abstract_inverted_index.text	40
abstract_inverted_index.that	34
abstract_inverted_index.well	67
abstract_inverted_index.with	57, 108
abstract_inverted_index.word	110
abstract_inverted_index.(WER)	113
abstract_inverted_index.0.73.	136
abstract_inverted_index.5.1%,	115
abstract_inverted_index.9.4%.	121
abstract_inverted_index.Dutch	104
abstract_inverted_index.aloud	2
abstract_inverted_index.audio	36
abstract_inverted_index.child	63, 105
abstract_inverted_index.error	111
abstract_inverted_index.large	53
abstract_inverted_index.model	92
abstract_inverted_index.novel	31
abstract_inverted_index.score	132
abstract_inverted_index.their	69
abstract_inverted_index.using	49
abstract_inverted_index.(LLMs)	56
abstract_inverted_index.models	55
abstract_inverted_index.speech	64
abstract_inverted_index.system	98
abstract_inverted_index.Whisper	50, 83, 91
abstract_inverted_index.improve	60
abstract_inverted_index.mistake	74, 127
abstract_inverted_index.present	29
abstract_inverted_index.prompts	58
abstract_inverted_index.provide	5
abstract_inverted_index.reading	1, 16, 21, 73, 126
abstract_inverted_index.remains	26
abstract_inverted_index.results	77
abstract_inverted_index.scoring	14
abstract_inverted_index.speech,	107
abstract_inverted_index.support	7
abstract_inverted_index.systems	23
abstract_inverted_index.without	93
abstract_inverted_index.However,	18
abstract_inverted_index.achieved	99
abstract_inverted_index.approach	33
abstract_inverted_index.baseline	90, 118
abstract_inverted_index.compared	87
abstract_inverted_index.enabling	11
abstract_inverted_index.explored	45
abstract_inverted_index.improved	125
abstract_inverted_index.language	54
abstract_inverted_index.limited.	27
abstract_inverted_index.research	19
abstract_inverted_index.teachers	9
abstract_inverted_index.valuable	6
abstract_inverted_index.Automatic	0
abstract_inverted_index.efficient	13
abstract_inverted_index.improving	116
abstract_inverted_index.knowledge	38
abstract_inverted_index.leverages	35
abstract_inverted_index.potential	47
abstract_inverted_index.prompting	82, 85
abstract_inverted_index.detection,	128
abstract_inverted_index.detection.	75
abstract_inverted_index.downstream	72
abstract_inverted_index.evaluation	3, 22
abstract_inverted_index.exercises.	17
abstract_inverted_index.increasing	129
abstract_inverted_index.multimodal	32
abstract_inverted_index.performing	97
abstract_inverted_index.prompting.	94
abstract_inverted_index.resources.	41
abstract_inverted_index.demonstrate	78
abstract_inverted_index.particular,	43
abstract_inverted_index.performance	102
abstract_inverted_index.recognition	101
abstract_inverted_index.Furthermore,	122
abstract_inverted_index.applications	25
abstract_inverted_index.recognition,	65
abstract_inverted_index.effectiveness	70, 80
abstract_inverted_index.significantly	124
abstract_inverted_index.transcriptions	61
abstract_inverted_index.state-of-the-art	100
abstract_inverted_index.instruction-tuned	52
cited_by_percentile_year
countries_distinct_count	0
institutions_distinct_count	4
citation_normalized_percentile.value	0.23208804
citation_normalized_percentile.is_in_top_1_percent	False
citation_normalized_percentile.is_in_top_10_percent	True