Data selection for speech recognition

This paper presents a strategy for efficiently selecting informative data from large corpora of transcribed speech. We propose to choose data uniformly according to the distribution of some target speech unit (phoneme, word, character, etc). In our experiment, in contrast to the common belief that "there is no data like more data", we found it possible to select a highly informative subset of data that produces recognition performance comparable to a system that makes use of a much larger amount of data. At the same time, our selection process is efficient and fast.

• Publication year

  2007

• Venue

  Automatic Speech Recognition & Understanding

• Publication date

  2007-12-01

• Fields of study

  Computer Science

• Identifiers
  DOI 10.1109/ASRU.2007.4430173
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• A New Data Selection Approach for Semi-Supervised Acoustic Modeling

  2006cited by this paper
• Challenges with Rapid Adaptation of Speech Translation Systems to New Language Pairs

  2006cited by this paper
• Unsupervised training of acoustic models for large vocabulary continuous speech recognition

  2005cited by this paper
• Active learning for spoken language understanding

  2003cited by this paper
• In search of optimal data selection for training of automatic speech recognition systems

  2003cited by this paper
• 私のすすめるこの一冊 ; Spoken Launguage Processing: A Guide to Theory, Algorithm, and System Development, Xuedong Huang, Alex Acero and Hsiao-Wuen Hon, Prentice Hall, 2001 年

  2003cited by this paper
• Active and unsupervised learning for automatic speech recognition

  2003cited by this paper
• Selective sampling of training data for speech recognition

  2002cited by this paper
• Lightly supervised and unsupervised acoustic model training

  2002cited by this paper
• Active learning for automatic speech recognition

  2002cited by this paper
• Optimal selection of speech data for automatic speech recognition systems

  2002cited by this paper
• The Elements of Statistical Learning: Data Mining, Inference, and Prediction

  2001cited by this paper
• Spoken Language Processing

  2001cited by this paper
• Pattern classification, 2nd Edition

  2000cited by this paper
• On large-vocabulary speaker-independent continuous speech recognition

  1988cited by this paper
• Large-vocabulary speaker-independent continuous speech recognition: the sphinx system

  1988cited by this paper

• An Exhaustive Evaluation of TTS- and VC-based Data Augmentation for ASR

  2025influential citation
• A Survey on Data Selection for Efficient Speech Processing

  2025cites this paper
• Combining X-Vectors and Bayesian Batch Active Learning: Two-Stage Active Learning Pipeline for Speech Recognition

  2024cites this paper
• Exploring emergent syllables in end-to-end automatic speech recognizers through model explainability technique

  2024cites this paper
• Speech Corpora Divergence Based Unsupervised Data Selection for ASR

  2023cites this paper
• Soft Random Sampling: A Theoretical and Empirical Analysis

  2023cites this paper
• Near-Optimal Active Learning for Multilingual Grapheme-to-Phoneme Conversion

  2023cites this paper
• Self-Supervised Dataset Pruning for Efficient Training in Audio Anti-spoofing

  2023cites this paper
• Targeted Subset Selection for Limited-data ASR Accent Adaptation

  2022cites this paper
• Representative Subset Selection for Efficient Fine-Tuning in Self-Supervised Speech Recognition

  2022cites this paper
• Speech Synthesis from Found Data

  2022cites this paper
• Towards Representative Subset Selection for Self-Supervised Speech Recognition

  2022cites this paper
• Dataset Pruning for Resource-constrained Spoofed Audio Detection

  2022cites this paper
• A Comparative Study of Bot Detection Techniques With an Application in Twitter Covid-19 Discourse

  2022cites this paper
• Effect of Training Data Selection for Speech Recognition of Emotional Speech

  2021cites this paper
• A comparative study of Bot Detection techniques methods with an application related to Covid-19 discourse on Twitter

  2021cites this paper
• Advanced Data Expoitation in Speech Analysis

  2021cites this paper
• DITTO: Data-efficient and Fair Targeted Subset Selection for ASR Accent Adaptation

  2021cites this paper
• Error-Driven Fixed-Budget ASR Personalization for Accented Speakers

  2021cites this paper
• Scalable and Generalizable Social Bot Detection through Data Selection

  2019cites this paper
• Modelling Sample Informativeness for Deep Affective Computing

  2019cites this paper
• Automatic Selection of Speech Data based on Confidence Measure

  2019cites this paper
• Semi-supervised and Active-learning Scenarios: Efficient Acoustic Model Refinement for a Low Resource Indian Language

  2018cites this paper
• Submodular Based Unsupervised Data Selection

  2018cites this paper
• Automated cleft speech evaluation using speech recognition.

  2017cites this paper
• Low-resource spoken keyword search strategies in georgian inspired by distinctive feature theory

  2017cites this paper
• Advanced Data Exploitation in Speech Analysis: An overview

  2017cites this paper
• Training Data Augmentation and Data Selection

  2017cites this paper
• Review of various stages in speaker recognition system, performance measures and recognition toolkits

  2017cites this paper
• Methods for addressing data diversity in automatic speech recognition

  2017influential citation
• Data Selection by Sequence Summarizing Neural Network in Mismatch Condition Training

  2016cites this paper
• SELECTION FOR NOISE ROBUST EXEMPLAR MATCHING

  2016cites this paper
• Cross-lingual deep neural network based submodular unbiased data selection for low-resource keyword search

  2016cites this paper
• Data selection for noise robust exemplar matching

  2016cites this paper
• A Comparative Study of the Performance of HMM, DNN, and RNN based Speech Synthesis Systems Trained on Very Large Speaker-Dependent Corpora

  2016cites this paper
• A fast query-by-example spoken term detection for zero resource languages

  2016cites this paper
• Acoustic modeling using state projection vectors of subspace Gaussian mixture model to train deep neural network on entropy maximized Hindi dataset

  2016cites this paper
• Data-selective transfer learning for multi-domain speech recognition

  2015cites this paper
• Multilingual representations for low resource speech recognition and keyword search

  2015cites this paper
• Low-resource keyword search strategies for tamil

  2015cites this paper
• Unsupervised data selection and word-morph mixed language model for tamil low-resource keyword search

  2015cites this paper
• A Multi-criteria Text Selection Approach for Building a Speech Corpus

  2015cites this paper
• Submodular data selection with acoustic and phonetic features for automatic speech recognition

  2015cites this paper
• Optimizing Data Selection for Automatic Speech Recognition in Low Resource Languages

  2015cites this paper
• Active learning based data selection for limited resource STT and KWS

  2015cites this paper
• Improving data selection for low-resource STT and KWS

  2015cites this paper
• An Agreement and Sparseness-based Learning Instance Selection and its Application to Subjective Speech Phenomena

  2014cites this paper
• PAKDD’12 best paper: generating balanced classifier-independent training samples from unlabeled data

  2014cites this paper
• Automatic selection of speakers for improved acoustic modelling: recognition of disordered speech with sparse data

  2014cites this paper
• Efficient data selection for speech recognition based on prior confidence estimation using speech and monophone models

  2014cites this paper
• Fast Multi-Stage Submodular Maximization : Extended version

  2014cites this paper
• Fast Multi-Stage Submodular Maximization : Extended version

  2014cites this paper
• UNSUPERVISED SUBMODULAR SUBSET SELECTION FOR SPEECH DATA : EXTENDED VERSION

  2014influential citation
• Efficient data selection for ASR

  2014influential citation
• Fast Multi-stage Submodular Maximization

  2014cites this paper
• Unsupervised submodular subset selection for speech data

  2014influential citation
• Submodular subset selection for large-scale speech training data

  2014cites this paper
• A submodular optimization approach to sentence set selection

  2014cites this paper
• Efficient use of training data for sinhala speech recognition using active learning

  2013cites this paper
• Using Document Summarization Techniques for Speech Data Subset Selection

  2013influential citation
• Automatic speech recognition for resource–scarce environments

  2013influential citation
• A Submodularity Framework for Data Subset Selection

  2013influential citation
• Towards natural speech acquisition: incremental word learning with limited data

  2013cites this paper
• Incremental word learning: Efficient HMM initialization and large margin discriminative adaptation

  2012cites this paper
• PAKDD’12 best paper: generating balanced classifier-independent training samples from unlabeled data

  2012cites this paper
• Collecting and evaluating speech recognition corpora for 11 South African languages

  2011cites this paper
• Age and gender detection in the I-DASH project

  2011cites this paper
• Kullback-Leibler Divergence-Based ASR Training Data Selection

  2011cites this paper
• Efficient data selection for speech recognition based on prior confidence estimation using speech and context independent models

  2011cites this paper
• Data Balancing for Efficient Training of Hybrid ANN/HMM Automatic Speech Recognition Systems

  2011cites this paper
• Brazilian portuguese acoustic model training based on data borrowing from other language

  2010cites this paper
• Data pruning for template-based automatic speech recognition

  2010cites this paper
• Dynamic language modeling for European Portuguese

  2010cites this paper
• Efficient data selection for spoken document retrieval based on prior confidence estimation using speech and context independent models

  2010cites this paper
• ASR corpus design for resource-scarce languages

  2009cites this paper
• How to select a good training-data subset for transcription: submodular active selection for sequences

  2009cites this paper
• Collecting and Evaluating Speech Recognition Corpora for Nine Southern Bantu Languages

  2009cites this paper
• Data sufficiency analysis for automatic speech recognition

  2009cites this paper
• Data requirements for speaker independent acoustic models

  2008cites this paper